Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2751 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2751 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 25.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3875,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.06451612903225806,
|
| 13 |
+
"grad_norm": 14.686903953552246,
|
| 14 |
+
"learning_rate": 5.154639175257732e-06,
|
| 15 |
+
"loss": 1.8113,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.12903225806451613,
|
| 20 |
+
"grad_norm": 9.743651390075684,
|
| 21 |
+
"learning_rate": 1.0309278350515464e-05,
|
| 22 |
+
"loss": 1.57,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.1935483870967742,
|
| 27 |
+
"grad_norm": 2.2041146755218506,
|
| 28 |
+
"learning_rate": 1.5463917525773197e-05,
|
| 29 |
+
"loss": 0.6814,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.25806451612903225,
|
| 34 |
+
"grad_norm": 3.186547040939331,
|
| 35 |
+
"learning_rate": 2.0618556701030927e-05,
|
| 36 |
+
"loss": 0.4245,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.3225806451612903,
|
| 41 |
+
"grad_norm": 1.2101554870605469,
|
| 42 |
+
"learning_rate": 2.5773195876288658e-05,
|
| 43 |
+
"loss": 0.2726,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.3870967741935484,
|
| 48 |
+
"grad_norm": 0.7474563121795654,
|
| 49 |
+
"learning_rate": 3.0927835051546395e-05,
|
| 50 |
+
"loss": 0.2088,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.45161290322580644,
|
| 55 |
+
"grad_norm": 0.9382892847061157,
|
| 56 |
+
"learning_rate": 3.6082474226804125e-05,
|
| 57 |
+
"loss": 0.2013,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.5161290322580645,
|
| 62 |
+
"grad_norm": 1.0286656618118286,
|
| 63 |
+
"learning_rate": 4.1237113402061855e-05,
|
| 64 |
+
"loss": 0.1683,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.5806451612903226,
|
| 69 |
+
"grad_norm": 0.867642343044281,
|
| 70 |
+
"learning_rate": 4.639175257731959e-05,
|
| 71 |
+
"loss": 0.1561,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.6451612903225806,
|
| 76 |
+
"grad_norm": 1.0434398651123047,
|
| 77 |
+
"learning_rate": 5.1546391752577315e-05,
|
| 78 |
+
"loss": 0.1509,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.7096774193548387,
|
| 83 |
+
"grad_norm": 1.3080216646194458,
|
| 84 |
+
"learning_rate": 5.670103092783505e-05,
|
| 85 |
+
"loss": 0.1291,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.7741935483870968,
|
| 90 |
+
"grad_norm": 0.38779398798942566,
|
| 91 |
+
"learning_rate": 6.185567010309279e-05,
|
| 92 |
+
"loss": 0.1209,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.8387096774193549,
|
| 97 |
+
"grad_norm": 0.49606090784072876,
|
| 98 |
+
"learning_rate": 6.701030927835051e-05,
|
| 99 |
+
"loss": 0.1202,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.9032258064516129,
|
| 104 |
+
"grad_norm": 0.7980489730834961,
|
| 105 |
+
"learning_rate": 7.216494845360825e-05,
|
| 106 |
+
"loss": 0.1153,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.967741935483871,
|
| 111 |
+
"grad_norm": 0.8646230697631836,
|
| 112 |
+
"learning_rate": 7.731958762886599e-05,
|
| 113 |
+
"loss": 0.1068,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 1.032258064516129,
|
| 118 |
+
"grad_norm": 0.6217169165611267,
|
| 119 |
+
"learning_rate": 8.247422680412371e-05,
|
| 120 |
+
"loss": 0.1049,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 1.096774193548387,
|
| 125 |
+
"grad_norm": 0.7635177373886108,
|
| 126 |
+
"learning_rate": 8.762886597938145e-05,
|
| 127 |
+
"loss": 0.0953,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.1612903225806452,
|
| 132 |
+
"grad_norm": 0.5218844413757324,
|
| 133 |
+
"learning_rate": 9.278350515463918e-05,
|
| 134 |
+
"loss": 0.098,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 1.2258064516129032,
|
| 139 |
+
"grad_norm": 0.5090600848197937,
|
| 140 |
+
"learning_rate": 9.793814432989691e-05,
|
| 141 |
+
"loss": 0.1049,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 1.2903225806451613,
|
| 146 |
+
"grad_norm": 0.5775400400161743,
|
| 147 |
+
"learning_rate": 9.999934444418978e-05,
|
| 148 |
+
"loss": 0.0995,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 1.3548387096774195,
|
| 153 |
+
"grad_norm": 0.739729106426239,
|
| 154 |
+
"learning_rate": 9.999533833204613e-05,
|
| 155 |
+
"loss": 0.0879,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 1.4193548387096775,
|
| 160 |
+
"grad_norm": 0.4800107479095459,
|
| 161 |
+
"learning_rate": 9.99876905968797e-05,
|
| 162 |
+
"loss": 0.0854,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 1.4838709677419355,
|
| 167 |
+
"grad_norm": 0.7302453517913818,
|
| 168 |
+
"learning_rate": 9.997640179574574e-05,
|
| 169 |
+
"loss": 0.0923,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.5483870967741935,
|
| 174 |
+
"grad_norm": 0.362665057182312,
|
| 175 |
+
"learning_rate": 9.996147275091215e-05,
|
| 176 |
+
"loss": 0.0873,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.6129032258064515,
|
| 181 |
+
"grad_norm": 0.502337634563446,
|
| 182 |
+
"learning_rate": 9.994290454979931e-05,
|
| 183 |
+
"loss": 0.0833,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.6774193548387095,
|
| 188 |
+
"grad_norm": 0.6111910939216614,
|
| 189 |
+
"learning_rate": 9.992069854490114e-05,
|
| 190 |
+
"loss": 0.0806,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 1.7419354838709677,
|
| 195 |
+
"grad_norm": 0.5922254920005798,
|
| 196 |
+
"learning_rate": 9.989485635368641e-05,
|
| 197 |
+
"loss": 0.0838,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.8064516129032258,
|
| 202 |
+
"grad_norm": 1.0255498886108398,
|
| 203 |
+
"learning_rate": 9.986537985848097e-05,
|
| 204 |
+
"loss": 0.0838,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 1.870967741935484,
|
| 209 |
+
"grad_norm": 0.8610944151878357,
|
| 210 |
+
"learning_rate": 9.983227120633066e-05,
|
| 211 |
+
"loss": 0.0831,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.935483870967742,
|
| 216 |
+
"grad_norm": 0.5282477736473083,
|
| 217 |
+
"learning_rate": 9.979553280884494e-05,
|
| 218 |
+
"loss": 0.0768,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 2.0,
|
| 223 |
+
"grad_norm": 0.7257622480392456,
|
| 224 |
+
"learning_rate": 9.975516734202118e-05,
|
| 225 |
+
"loss": 0.082,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 2.064516129032258,
|
| 230 |
+
"grad_norm": 0.5951042771339417,
|
| 231 |
+
"learning_rate": 9.971117774604977e-05,
|
| 232 |
+
"loss": 0.0785,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 2.129032258064516,
|
| 237 |
+
"grad_norm": 1.008421778678894,
|
| 238 |
+
"learning_rate": 9.966356722509995e-05,
|
| 239 |
+
"loss": 0.0756,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 2.193548387096774,
|
| 244 |
+
"grad_norm": 0.6158808469772339,
|
| 245 |
+
"learning_rate": 9.961233924708644e-05,
|
| 246 |
+
"loss": 0.0698,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 2.258064516129032,
|
| 251 |
+
"grad_norm": 0.5757787227630615,
|
| 252 |
+
"learning_rate": 9.95574975434168e-05,
|
| 253 |
+
"loss": 0.0703,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 2.3225806451612905,
|
| 258 |
+
"grad_norm": 0.6450510621070862,
|
| 259 |
+
"learning_rate": 9.949904610871968e-05,
|
| 260 |
+
"loss": 0.0685,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 2.3870967741935485,
|
| 265 |
+
"grad_norm": 0.5834465026855469,
|
| 266 |
+
"learning_rate": 9.943698920055384e-05,
|
| 267 |
+
"loss": 0.0676,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 2.4516129032258065,
|
| 272 |
+
"grad_norm": 0.634287416934967,
|
| 273 |
+
"learning_rate": 9.937133133909801e-05,
|
| 274 |
+
"loss": 0.0639,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 2.5161290322580645,
|
| 279 |
+
"grad_norm": 0.32366201281547546,
|
| 280 |
+
"learning_rate": 9.93020773068217e-05,
|
| 281 |
+
"loss": 0.0734,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 2.5806451612903225,
|
| 286 |
+
"grad_norm": 0.6290192008018494,
|
| 287 |
+
"learning_rate": 9.922923214813675e-05,
|
| 288 |
+
"loss": 0.0748,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 2.6451612903225805,
|
| 293 |
+
"grad_norm": 0.5695043206214905,
|
| 294 |
+
"learning_rate": 9.915280116903003e-05,
|
| 295 |
+
"loss": 0.066,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 2.709677419354839,
|
| 300 |
+
"grad_norm": 0.5956539511680603,
|
| 301 |
+
"learning_rate": 9.907278993667681e-05,
|
| 302 |
+
"loss": 0.065,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 2.774193548387097,
|
| 307 |
+
"grad_norm": 0.505896806716919,
|
| 308 |
+
"learning_rate": 9.89892042790354e-05,
|
| 309 |
+
"loss": 0.0637,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 2.838709677419355,
|
| 314 |
+
"grad_norm": 0.6462914943695068,
|
| 315 |
+
"learning_rate": 9.890205028442252e-05,
|
| 316 |
+
"loss": 0.0742,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 2.903225806451613,
|
| 321 |
+
"grad_norm": 0.6515319347381592,
|
| 322 |
+
"learning_rate": 9.88113343010699e-05,
|
| 323 |
+
"loss": 0.068,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 2.967741935483871,
|
| 328 |
+
"grad_norm": 0.5437873601913452,
|
| 329 |
+
"learning_rate": 9.87170629366619e-05,
|
| 330 |
+
"loss": 0.0734,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 3.032258064516129,
|
| 335 |
+
"grad_norm": 0.5710848569869995,
|
| 336 |
+
"learning_rate": 9.861924305785409e-05,
|
| 337 |
+
"loss": 0.072,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 3.096774193548387,
|
| 342 |
+
"grad_norm": 0.47411492466926575,
|
| 343 |
+
"learning_rate": 9.851788178977327e-05,
|
| 344 |
+
"loss": 0.0632,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 3.161290322580645,
|
| 349 |
+
"grad_norm": 0.40361443161964417,
|
| 350 |
+
"learning_rate": 9.841298651549831e-05,
|
| 351 |
+
"loss": 0.058,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 3.225806451612903,
|
| 356 |
+
"grad_norm": 0.5025441646575928,
|
| 357 |
+
"learning_rate": 9.83045648755225e-05,
|
| 358 |
+
"loss": 0.0594,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 3.2903225806451615,
|
| 363 |
+
"grad_norm": 0.6928324103355408,
|
| 364 |
+
"learning_rate": 9.819262476719692e-05,
|
| 365 |
+
"loss": 0.061,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 3.3548387096774195,
|
| 370 |
+
"grad_norm": 0.45038920640945435,
|
| 371 |
+
"learning_rate": 9.807717434415527e-05,
|
| 372 |
+
"loss": 0.0597,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 3.4193548387096775,
|
| 377 |
+
"grad_norm": 0.35843929648399353,
|
| 378 |
+
"learning_rate": 9.795822201571998e-05,
|
| 379 |
+
"loss": 0.0637,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 3.4838709677419355,
|
| 384 |
+
"grad_norm": 0.727299690246582,
|
| 385 |
+
"learning_rate": 9.78357764462896e-05,
|
| 386 |
+
"loss": 0.062,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 3.5483870967741935,
|
| 391 |
+
"grad_norm": 0.5294666886329651,
|
| 392 |
+
"learning_rate": 9.770984655470775e-05,
|
| 393 |
+
"loss": 0.0616,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 3.6129032258064515,
|
| 398 |
+
"grad_norm": 0.24387721717357635,
|
| 399 |
+
"learning_rate": 9.758044151361347e-05,
|
| 400 |
+
"loss": 0.0599,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 3.6774193548387095,
|
| 405 |
+
"grad_norm": 0.42391934990882874,
|
| 406 |
+
"learning_rate": 9.744757074877312e-05,
|
| 407 |
+
"loss": 0.0572,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 3.741935483870968,
|
| 412 |
+
"grad_norm": 0.43955346941947937,
|
| 413 |
+
"learning_rate": 9.731124393839375e-05,
|
| 414 |
+
"loss": 0.0562,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 3.806451612903226,
|
| 419 |
+
"grad_norm": 0.6420799493789673,
|
| 420 |
+
"learning_rate": 9.717147101241816e-05,
|
| 421 |
+
"loss": 0.0568,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 3.870967741935484,
|
| 426 |
+
"grad_norm": 0.2555811405181885,
|
| 427 |
+
"learning_rate": 9.70282621518017e-05,
|
| 428 |
+
"loss": 0.0539,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 3.935483870967742,
|
| 433 |
+
"grad_norm": 0.408918559551239,
|
| 434 |
+
"learning_rate": 9.688162778777056e-05,
|
| 435 |
+
"loss": 0.0585,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 4.0,
|
| 440 |
+
"grad_norm": 0.422941118478775,
|
| 441 |
+
"learning_rate": 9.673157860106206e-05,
|
| 442 |
+
"loss": 0.0547,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 4.064516129032258,
|
| 447 |
+
"grad_norm": 0.4911976158618927,
|
| 448 |
+
"learning_rate": 9.657812552114667e-05,
|
| 449 |
+
"loss": 0.0559,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 4.129032258064516,
|
| 454 |
+
"grad_norm": 0.715604841709137,
|
| 455 |
+
"learning_rate": 9.642127972543184e-05,
|
| 456 |
+
"loss": 0.0536,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 4.193548387096774,
|
| 461 |
+
"grad_norm": 0.530028760433197,
|
| 462 |
+
"learning_rate": 9.626105263844797e-05,
|
| 463 |
+
"loss": 0.0538,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 4.258064516129032,
|
| 468 |
+
"grad_norm": 0.4460236728191376,
|
| 469 |
+
"learning_rate": 9.609745593101611e-05,
|
| 470 |
+
"loss": 0.0533,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 4.32258064516129,
|
| 475 |
+
"grad_norm": 0.422241747379303,
|
| 476 |
+
"learning_rate": 9.5930501519398e-05,
|
| 477 |
+
"loss": 0.0573,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 4.387096774193548,
|
| 482 |
+
"grad_norm": 0.5292857885360718,
|
| 483 |
+
"learning_rate": 9.576020156442802e-05,
|
| 484 |
+
"loss": 0.0558,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 4.451612903225806,
|
| 489 |
+
"grad_norm": 0.3689402937889099,
|
| 490 |
+
"learning_rate": 9.55865684706274e-05,
|
| 491 |
+
"loss": 0.0502,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 4.516129032258064,
|
| 496 |
+
"grad_norm": 0.490133672952652,
|
| 497 |
+
"learning_rate": 9.540961488530074e-05,
|
| 498 |
+
"loss": 0.0574,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 4.580645161290323,
|
| 503 |
+
"grad_norm": 0.40120774507522583,
|
| 504 |
+
"learning_rate": 9.522935369761475e-05,
|
| 505 |
+
"loss": 0.0557,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 4.645161290322581,
|
| 510 |
+
"grad_norm": 0.4087425172328949,
|
| 511 |
+
"learning_rate": 9.504579803765943e-05,
|
| 512 |
+
"loss": 0.0504,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 4.709677419354839,
|
| 517 |
+
"grad_norm": 0.6458566188812256,
|
| 518 |
+
"learning_rate": 9.485896127549163e-05,
|
| 519 |
+
"loss": 0.0564,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 4.774193548387097,
|
| 524 |
+
"grad_norm": 0.43900686502456665,
|
| 525 |
+
"learning_rate": 9.466885702016127e-05,
|
| 526 |
+
"loss": 0.0566,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 4.838709677419355,
|
| 531 |
+
"grad_norm": 0.4088176190853119,
|
| 532 |
+
"learning_rate": 9.447549911872002e-05,
|
| 533 |
+
"loss": 0.058,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 4.903225806451613,
|
| 538 |
+
"grad_norm": 0.40604203939437866,
|
| 539 |
+
"learning_rate": 9.427890165521269e-05,
|
| 540 |
+
"loss": 0.0499,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 4.967741935483871,
|
| 545 |
+
"grad_norm": 0.41275304555892944,
|
| 546 |
+
"learning_rate": 9.407907894965137e-05,
|
| 547 |
+
"loss": 0.0499,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 5.032258064516129,
|
| 552 |
+
"grad_norm": 0.2919212281703949,
|
| 553 |
+
"learning_rate": 9.387604555697239e-05,
|
| 554 |
+
"loss": 0.0555,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 5.096774193548387,
|
| 559 |
+
"grad_norm": 0.39005234837532043,
|
| 560 |
+
"learning_rate": 9.366981626597608e-05,
|
| 561 |
+
"loss": 0.0502,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 5.161290322580645,
|
| 566 |
+
"grad_norm": 0.4438292980194092,
|
| 567 |
+
"learning_rate": 9.346040609824962e-05,
|
| 568 |
+
"loss": 0.0568,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 5.225806451612903,
|
| 573 |
+
"grad_norm": 0.5161975622177124,
|
| 574 |
+
"learning_rate": 9.324783030707292e-05,
|
| 575 |
+
"loss": 0.0532,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 5.290322580645161,
|
| 580 |
+
"grad_norm": 0.40832287073135376,
|
| 581 |
+
"learning_rate": 9.303210437630746e-05,
|
| 582 |
+
"loss": 0.0552,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 5.354838709677419,
|
| 587 |
+
"grad_norm": 0.3923545777797699,
|
| 588 |
+
"learning_rate": 9.28132440192686e-05,
|
| 589 |
+
"loss": 0.0557,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 5.419354838709677,
|
| 594 |
+
"grad_norm": 0.33239755034446716,
|
| 595 |
+
"learning_rate": 9.25912651775809e-05,
|
| 596 |
+
"loss": 0.054,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 5.483870967741936,
|
| 601 |
+
"grad_norm": 0.40319114923477173,
|
| 602 |
+
"learning_rate": 9.236618402001707e-05,
|
| 603 |
+
"loss": 0.0501,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 5.548387096774194,
|
| 608 |
+
"grad_norm": 0.49930399656295776,
|
| 609 |
+
"learning_rate": 9.213801694132012e-05,
|
| 610 |
+
"loss": 0.0541,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 5.612903225806452,
|
| 615 |
+
"grad_norm": 0.3818790912628174,
|
| 616 |
+
"learning_rate": 9.190678056100933e-05,
|
| 617 |
+
"loss": 0.0501,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 5.67741935483871,
|
| 622 |
+
"grad_norm": 0.5576165914535522,
|
| 623 |
+
"learning_rate": 9.16724917221695e-05,
|
| 624 |
+
"loss": 0.0506,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 5.741935483870968,
|
| 629 |
+
"grad_norm": 0.5411040782928467,
|
| 630 |
+
"learning_rate": 9.143516749022433e-05,
|
| 631 |
+
"loss": 0.0461,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 5.806451612903226,
|
| 636 |
+
"grad_norm": 0.35116180777549744,
|
| 637 |
+
"learning_rate": 9.119482515169324e-05,
|
| 638 |
+
"loss": 0.0479,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 5.870967741935484,
|
| 643 |
+
"grad_norm": 0.30685099959373474,
|
| 644 |
+
"learning_rate": 9.095148221293225e-05,
|
| 645 |
+
"loss": 0.0459,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 5.935483870967742,
|
| 650 |
+
"grad_norm": 0.4129478633403778,
|
| 651 |
+
"learning_rate": 9.070515639885888e-05,
|
| 652 |
+
"loss": 0.0505,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 6.0,
|
| 657 |
+
"grad_norm": 0.5053979754447937,
|
| 658 |
+
"learning_rate": 9.045586565166107e-05,
|
| 659 |
+
"loss": 0.0525,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 6.064516129032258,
|
| 664 |
+
"grad_norm": 0.3175376057624817,
|
| 665 |
+
"learning_rate": 9.020362812949024e-05,
|
| 666 |
+
"loss": 0.0456,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 6.129032258064516,
|
| 671 |
+
"grad_norm": 0.4626881182193756,
|
| 672 |
+
"learning_rate": 8.994846220513872e-05,
|
| 673 |
+
"loss": 0.0441,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 6.193548387096774,
|
| 678 |
+
"grad_norm": 0.321857750415802,
|
| 679 |
+
"learning_rate": 8.969038646470143e-05,
|
| 680 |
+
"loss": 0.0468,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 6.258064516129032,
|
| 685 |
+
"grad_norm": 0.4286996126174927,
|
| 686 |
+
"learning_rate": 8.942941970622216e-05,
|
| 687 |
+
"loss": 0.0488,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 6.32258064516129,
|
| 692 |
+
"grad_norm": 0.5266990065574646,
|
| 693 |
+
"learning_rate": 8.916558093832424e-05,
|
| 694 |
+
"loss": 0.0478,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 6.387096774193548,
|
| 699 |
+
"grad_norm": 0.2945270538330078,
|
| 700 |
+
"learning_rate": 8.889888937882604e-05,
|
| 701 |
+
"loss": 0.052,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 6.451612903225806,
|
| 706 |
+
"grad_norm": 0.32375168800354004,
|
| 707 |
+
"learning_rate": 8.862936445334116e-05,
|
| 708 |
+
"loss": 0.0484,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 6.516129032258064,
|
| 713 |
+
"grad_norm": 0.3967803120613098,
|
| 714 |
+
"learning_rate": 8.835702579386342e-05,
|
| 715 |
+
"loss": 0.0549,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 6.580645161290323,
|
| 720 |
+
"grad_norm": 0.35367152094841003,
|
| 721 |
+
"learning_rate": 8.808189323733693e-05,
|
| 722 |
+
"loss": 0.0476,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 6.645161290322581,
|
| 727 |
+
"grad_norm": 0.46086496114730835,
|
| 728 |
+
"learning_rate": 8.780398682421117e-05,
|
| 729 |
+
"loss": 0.0463,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 6.709677419354839,
|
| 734 |
+
"grad_norm": 0.5151063799858093,
|
| 735 |
+
"learning_rate": 8.752332679698128e-05,
|
| 736 |
+
"loss": 0.0468,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 6.774193548387097,
|
| 741 |
+
"grad_norm": 0.5349656343460083,
|
| 742 |
+
"learning_rate": 8.723993359871354e-05,
|
| 743 |
+
"loss": 0.0472,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 6.838709677419355,
|
| 748 |
+
"grad_norm": 0.4370066821575165,
|
| 749 |
+
"learning_rate": 8.695382787155643e-05,
|
| 750 |
+
"loss": 0.0414,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 6.903225806451613,
|
| 755 |
+
"grad_norm": 0.39678850769996643,
|
| 756 |
+
"learning_rate": 8.666503045523692e-05,
|
| 757 |
+
"loss": 0.0399,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 6.967741935483871,
|
| 762 |
+
"grad_norm": 0.5493939518928528,
|
| 763 |
+
"learning_rate": 8.63735623855427e-05,
|
| 764 |
+
"loss": 0.0545,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 7.032258064516129,
|
| 769 |
+
"grad_norm": 0.2903650104999542,
|
| 770 |
+
"learning_rate": 8.607944489278976e-05,
|
| 771 |
+
"loss": 0.0423,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 7.096774193548387,
|
| 776 |
+
"grad_norm": 0.5154908895492554,
|
| 777 |
+
"learning_rate": 8.578269940027616e-05,
|
| 778 |
+
"loss": 0.0441,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 7.161290322580645,
|
| 783 |
+
"grad_norm": 0.3061012625694275,
|
| 784 |
+
"learning_rate": 8.548334752272146e-05,
|
| 785 |
+
"loss": 0.0455,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 7.225806451612903,
|
| 790 |
+
"grad_norm": 0.34137821197509766,
|
| 791 |
+
"learning_rate": 8.518141106469233e-05,
|
| 792 |
+
"loss": 0.0426,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 7.290322580645161,
|
| 797 |
+
"grad_norm": 0.2501906156539917,
|
| 798 |
+
"learning_rate": 8.487691201901439e-05,
|
| 799 |
+
"loss": 0.0471,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 7.354838709677419,
|
| 804 |
+
"grad_norm": 0.5570168495178223,
|
| 805 |
+
"learning_rate": 8.456987256517023e-05,
|
| 806 |
+
"loss": 0.0415,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 7.419354838709677,
|
| 811 |
+
"grad_norm": 0.36885467171669006,
|
| 812 |
+
"learning_rate": 8.426031506768381e-05,
|
| 813 |
+
"loss": 0.0488,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 7.483870967741936,
|
| 818 |
+
"grad_norm": 0.2259949892759323,
|
| 819 |
+
"learning_rate": 8.394826207449157e-05,
|
| 820 |
+
"loss": 0.0443,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 7.548387096774194,
|
| 825 |
+
"grad_norm": 0.31414613127708435,
|
| 826 |
+
"learning_rate": 8.363373631529998e-05,
|
| 827 |
+
"loss": 0.0462,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 7.612903225806452,
|
| 832 |
+
"grad_norm": 0.43513432145118713,
|
| 833 |
+
"learning_rate": 8.331676069992989e-05,
|
| 834 |
+
"loss": 0.0417,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 7.67741935483871,
|
| 839 |
+
"grad_norm": 0.2809264361858368,
|
| 840 |
+
"learning_rate": 8.299735831664788e-05,
|
| 841 |
+
"loss": 0.0448,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 7.741935483870968,
|
| 846 |
+
"grad_norm": 0.3759617209434509,
|
| 847 |
+
"learning_rate": 8.267555243048446e-05,
|
| 848 |
+
"loss": 0.0432,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 7.806451612903226,
|
| 853 |
+
"grad_norm": 0.33138594031333923,
|
| 854 |
+
"learning_rate": 8.235136648153954e-05,
|
| 855 |
+
"loss": 0.0434,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 7.870967741935484,
|
| 860 |
+
"grad_norm": 0.47733554244041443,
|
| 861 |
+
"learning_rate": 8.202482408327497e-05,
|
| 862 |
+
"loss": 0.0406,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 7.935483870967742,
|
| 867 |
+
"grad_norm": 0.3868086338043213,
|
| 868 |
+
"learning_rate": 8.16959490207946e-05,
|
| 869 |
+
"loss": 0.0394,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 8.0,
|
| 874 |
+
"grad_norm": 0.44184964895248413,
|
| 875 |
+
"learning_rate": 8.136476524911187e-05,
|
| 876 |
+
"loss": 0.0393,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 8.064516129032258,
|
| 881 |
+
"grad_norm": 0.3758850693702698,
|
| 882 |
+
"learning_rate": 8.103129689140479e-05,
|
| 883 |
+
"loss": 0.0454,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 8.129032258064516,
|
| 888 |
+
"grad_norm": 0.5086485743522644,
|
| 889 |
+
"learning_rate": 8.069556823725895e-05,
|
| 890 |
+
"loss": 0.042,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 8.193548387096774,
|
| 895 |
+
"grad_norm": 0.3741207420825958,
|
| 896 |
+
"learning_rate": 8.03576037408982e-05,
|
| 897 |
+
"loss": 0.0439,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 8.258064516129032,
|
| 902 |
+
"grad_norm": 0.49306967854499817,
|
| 903 |
+
"learning_rate": 8.001742801940349e-05,
|
| 904 |
+
"loss": 0.04,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 8.32258064516129,
|
| 909 |
+
"grad_norm": 0.3498378098011017,
|
| 910 |
+
"learning_rate": 7.96750658509198e-05,
|
| 911 |
+
"loss": 0.0448,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 8.387096774193548,
|
| 916 |
+
"grad_norm": 0.3390824496746063,
|
| 917 |
+
"learning_rate": 7.933054217285118e-05,
|
| 918 |
+
"loss": 0.0472,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 8.451612903225806,
|
| 923 |
+
"grad_norm": 0.27651500701904297,
|
| 924 |
+
"learning_rate": 7.898388208004449e-05,
|
| 925 |
+
"loss": 0.0368,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 8.516129032258064,
|
| 930 |
+
"grad_norm": 0.33945196866989136,
|
| 931 |
+
"learning_rate": 7.86351108229614e-05,
|
| 932 |
+
"loss": 0.0376,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 8.580645161290322,
|
| 937 |
+
"grad_norm": 0.2656388580799103,
|
| 938 |
+
"learning_rate": 7.828425380583923e-05,
|
| 939 |
+
"loss": 0.0415,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 8.64516129032258,
|
| 944 |
+
"grad_norm": 0.2991732954978943,
|
| 945 |
+
"learning_rate": 7.793133658484051e-05,
|
| 946 |
+
"loss": 0.0434,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 8.709677419354838,
|
| 951 |
+
"grad_norm": 0.5004262328147888,
|
| 952 |
+
"learning_rate": 7.757638486619144e-05,
|
| 953 |
+
"loss": 0.0418,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 8.774193548387096,
|
| 958 |
+
"grad_norm": 0.3629301190376282,
|
| 959 |
+
"learning_rate": 7.72194245043095e-05,
|
| 960 |
+
"loss": 0.0368,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 8.838709677419354,
|
| 965 |
+
"grad_norm": 0.4232747256755829,
|
| 966 |
+
"learning_rate": 7.686048149992032e-05,
|
| 967 |
+
"loss": 0.0393,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 8.903225806451612,
|
| 972 |
+
"grad_norm": 0.3285137116909027,
|
| 973 |
+
"learning_rate": 7.649958199816363e-05,
|
| 974 |
+
"loss": 0.0373,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 8.967741935483872,
|
| 979 |
+
"grad_norm": 0.3300577402114868,
|
| 980 |
+
"learning_rate": 7.613675228668904e-05,
|
| 981 |
+
"loss": 0.0376,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 9.03225806451613,
|
| 986 |
+
"grad_norm": 0.27525898814201355,
|
| 987 |
+
"learning_rate": 7.577201879374115e-05,
|
| 988 |
+
"loss": 0.0401,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 9.096774193548388,
|
| 993 |
+
"grad_norm": 0.2236848622560501,
|
| 994 |
+
"learning_rate": 7.540540808623455e-05,
|
| 995 |
+
"loss": 0.0407,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 9.161290322580646,
|
| 1000 |
+
"grad_norm": 0.25309398770332336,
|
| 1001 |
+
"learning_rate": 7.50369468678188e-05,
|
| 1002 |
+
"loss": 0.0373,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 9.225806451612904,
|
| 1007 |
+
"grad_norm": 0.46527889370918274,
|
| 1008 |
+
"learning_rate": 7.466666197693326e-05,
|
| 1009 |
+
"loss": 0.0397,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 9.290322580645162,
|
| 1014 |
+
"grad_norm": 0.5017492175102234,
|
| 1015 |
+
"learning_rate": 7.429458038485222e-05,
|
| 1016 |
+
"loss": 0.0429,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 9.35483870967742,
|
| 1021 |
+
"grad_norm": 0.3527701795101166,
|
| 1022 |
+
"learning_rate": 7.392072919372032e-05,
|
| 1023 |
+
"loss": 0.0356,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 9.419354838709678,
|
| 1028 |
+
"grad_norm": 0.37153077125549316,
|
| 1029 |
+
"learning_rate": 7.354513563457855e-05,
|
| 1030 |
+
"loss": 0.038,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 9.483870967741936,
|
| 1035 |
+
"grad_norm": 0.5790444016456604,
|
| 1036 |
+
"learning_rate": 7.316782706538056e-05,
|
| 1037 |
+
"loss": 0.0408,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 9.548387096774194,
|
| 1042 |
+
"grad_norm": 0.3680723309516907,
|
| 1043 |
+
"learning_rate": 7.278883096900017e-05,
|
| 1044 |
+
"loss": 0.0402,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 9.612903225806452,
|
| 1049 |
+
"grad_norm": 0.40202105045318604,
|
| 1050 |
+
"learning_rate": 7.240817495122935e-05,
|
| 1051 |
+
"loss": 0.0376,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 9.67741935483871,
|
| 1056 |
+
"grad_norm": 0.24831123650074005,
|
| 1057 |
+
"learning_rate": 7.202588673876751e-05,
|
| 1058 |
+
"loss": 0.0371,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 9.741935483870968,
|
| 1063 |
+
"grad_norm": 0.4128989279270172,
|
| 1064 |
+
"learning_rate": 7.16419941772019e-05,
|
| 1065 |
+
"loss": 0.0337,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 9.806451612903226,
|
| 1070 |
+
"grad_norm": 0.2867458164691925,
|
| 1071 |
+
"learning_rate": 7.125652522897936e-05,
|
| 1072 |
+
"loss": 0.0383,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 9.870967741935484,
|
| 1077 |
+
"grad_norm": 0.32444578409194946,
|
| 1078 |
+
"learning_rate": 7.086950797136959e-05,
|
| 1079 |
+
"loss": 0.0425,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 9.935483870967742,
|
| 1084 |
+
"grad_norm": 0.8138065338134766,
|
| 1085 |
+
"learning_rate": 7.04809705944199e-05,
|
| 1086 |
+
"loss": 0.0422,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 10.0,
|
| 1091 |
+
"grad_norm": 0.5340340733528137,
|
| 1092 |
+
"learning_rate": 7.009094139890206e-05,
|
| 1093 |
+
"loss": 0.0386,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 10.064516129032258,
|
| 1098 |
+
"grad_norm": 0.3925679922103882,
|
| 1099 |
+
"learning_rate": 6.969944879425069e-05,
|
| 1100 |
+
"loss": 0.0375,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 10.129032258064516,
|
| 1105 |
+
"grad_norm": 0.32105499505996704,
|
| 1106 |
+
"learning_rate": 6.930652129649411e-05,
|
| 1107 |
+
"loss": 0.0412,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 10.193548387096774,
|
| 1112 |
+
"grad_norm": 0.31760817766189575,
|
| 1113 |
+
"learning_rate": 6.891218752617716e-05,
|
| 1114 |
+
"loss": 0.034,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 10.258064516129032,
|
| 1119 |
+
"grad_norm": 0.36178886890411377,
|
| 1120 |
+
"learning_rate": 6.851647620627648e-05,
|
| 1121 |
+
"loss": 0.0338,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 10.32258064516129,
|
| 1126 |
+
"grad_norm": 0.29000788927078247,
|
| 1127 |
+
"learning_rate": 6.811941616010847e-05,
|
| 1128 |
+
"loss": 0.0379,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 10.387096774193548,
|
| 1133 |
+
"grad_norm": 0.40512529015541077,
|
| 1134 |
+
"learning_rate": 6.772103630922964e-05,
|
| 1135 |
+
"loss": 0.0399,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 10.451612903225806,
|
| 1140 |
+
"grad_norm": 0.3559243083000183,
|
| 1141 |
+
"learning_rate": 6.732136567133019e-05,
|
| 1142 |
+
"loss": 0.0349,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 10.516129032258064,
|
| 1147 |
+
"grad_norm": 0.5191022157669067,
|
| 1148 |
+
"learning_rate": 6.69204333581202e-05,
|
| 1149 |
+
"loss": 0.0342,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 10.580645161290322,
|
| 1154 |
+
"grad_norm": 0.21767917275428772,
|
| 1155 |
+
"learning_rate": 6.651826857320926e-05,
|
| 1156 |
+
"loss": 0.0386,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 10.64516129032258,
|
| 1161 |
+
"grad_norm": 0.39933088421821594,
|
| 1162 |
+
"learning_rate": 6.611490060997933e-05,
|
| 1163 |
+
"loss": 0.0375,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 10.709677419354838,
|
| 1168 |
+
"grad_norm": 0.32508501410484314,
|
| 1169 |
+
"learning_rate": 6.57103588494509e-05,
|
| 1170 |
+
"loss": 0.0348,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 10.774193548387096,
|
| 1175 |
+
"grad_norm": 0.6718910932540894,
|
| 1176 |
+
"learning_rate": 6.530467275814301e-05,
|
| 1177 |
+
"loss": 0.0405,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 10.838709677419354,
|
| 1182 |
+
"grad_norm": 0.46869444847106934,
|
| 1183 |
+
"learning_rate": 6.489787188592693e-05,
|
| 1184 |
+
"loss": 0.035,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 10.903225806451612,
|
| 1189 |
+
"grad_norm": 0.39346209168434143,
|
| 1190 |
+
"learning_rate": 6.448998586387365e-05,
|
| 1191 |
+
"loss": 0.0344,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 10.967741935483872,
|
| 1196 |
+
"grad_norm": 0.4559718370437622,
|
| 1197 |
+
"learning_rate": 6.408104440209574e-05,
|
| 1198 |
+
"loss": 0.038,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 11.03225806451613,
|
| 1203 |
+
"grad_norm": 0.7514871954917908,
|
| 1204 |
+
"learning_rate": 6.367107728758319e-05,
|
| 1205 |
+
"loss": 0.0339,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 11.096774193548388,
|
| 1210 |
+
"grad_norm": 0.2884572148323059,
|
| 1211 |
+
"learning_rate": 6.326011438203378e-05,
|
| 1212 |
+
"loss": 0.0362,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 11.161290322580646,
|
| 1217 |
+
"grad_norm": 0.3102618157863617,
|
| 1218 |
+
"learning_rate": 6.28481856196779e-05,
|
| 1219 |
+
"loss": 0.0326,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 11.225806451612904,
|
| 1224 |
+
"grad_norm": 0.25867921113967896,
|
| 1225 |
+
"learning_rate": 6.24353210050983e-05,
|
| 1226 |
+
"loss": 0.0398,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 11.290322580645162,
|
| 1231 |
+
"grad_norm": 0.28058162331581116,
|
| 1232 |
+
"learning_rate": 6.202155061104448e-05,
|
| 1233 |
+
"loss": 0.0362,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 11.35483870967742,
|
| 1238 |
+
"grad_norm": 0.360380619764328,
|
| 1239 |
+
"learning_rate": 6.160690457624223e-05,
|
| 1240 |
+
"loss": 0.0374,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 11.419354838709678,
|
| 1245 |
+
"grad_norm": 0.40499451756477356,
|
| 1246 |
+
"learning_rate": 6.119141310319838e-05,
|
| 1247 |
+
"loss": 0.0355,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 11.483870967741936,
|
| 1252 |
+
"grad_norm": 0.3203999996185303,
|
| 1253 |
+
"learning_rate": 6.077510645600081e-05,
|
| 1254 |
+
"loss": 0.0349,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 11.548387096774194,
|
| 1259 |
+
"grad_norm": 0.2518363296985626,
|
| 1260 |
+
"learning_rate": 6.035801495811411e-05,
|
| 1261 |
+
"loss": 0.0356,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 11.612903225806452,
|
| 1266 |
+
"grad_norm": 0.4911201000213623,
|
| 1267 |
+
"learning_rate": 5.99401689901708e-05,
|
| 1268 |
+
"loss": 0.0331,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 11.67741935483871,
|
| 1273 |
+
"grad_norm": 0.30413028597831726,
|
| 1274 |
+
"learning_rate": 5.9521598987758395e-05,
|
| 1275 |
+
"loss": 0.0333,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 11.741935483870968,
|
| 1280 |
+
"grad_norm": 0.37464162707328796,
|
| 1281 |
+
"learning_rate": 5.910233543920263e-05,
|
| 1282 |
+
"loss": 0.0336,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 11.806451612903226,
|
| 1287 |
+
"grad_norm": 0.352731853723526,
|
| 1288 |
+
"learning_rate": 5.868240888334653e-05,
|
| 1289 |
+
"loss": 0.0327,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 11.870967741935484,
|
| 1294 |
+
"grad_norm": 0.4417053461074829,
|
| 1295 |
+
"learning_rate": 5.826184990732609e-05,
|
| 1296 |
+
"loss": 0.0327,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 11.935483870967742,
|
| 1301 |
+
"grad_norm": 0.3826858401298523,
|
| 1302 |
+
"learning_rate": 5.784068914434238e-05,
|
| 1303 |
+
"loss": 0.0333,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 12.0,
|
| 1308 |
+
"grad_norm": 0.4200211763381958,
|
| 1309 |
+
"learning_rate": 5.741895727143013e-05,
|
| 1310 |
+
"loss": 0.0368,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 12.064516129032258,
|
| 1315 |
+
"grad_norm": 0.22449392080307007,
|
| 1316 |
+
"learning_rate": 5.699668500722328e-05,
|
| 1317 |
+
"loss": 0.0326,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 12.129032258064516,
|
| 1322 |
+
"grad_norm": 0.2556811571121216,
|
| 1323 |
+
"learning_rate": 5.65739031097175e-05,
|
| 1324 |
+
"loss": 0.0305,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 12.193548387096774,
|
| 1329 |
+
"grad_norm": 0.34918874502182007,
|
| 1330 |
+
"learning_rate": 5.615064237402974e-05,
|
| 1331 |
+
"loss": 0.0294,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 12.258064516129032,
|
| 1336 |
+
"grad_norm": 0.3033507764339447,
|
| 1337 |
+
"learning_rate": 5.5726933630155196e-05,
|
| 1338 |
+
"loss": 0.0311,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 12.32258064516129,
|
| 1343 |
+
"grad_norm": 0.2633187770843506,
|
| 1344 |
+
"learning_rate": 5.5302807740721606e-05,
|
| 1345 |
+
"loss": 0.0369,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 12.387096774193548,
|
| 1350 |
+
"grad_norm": 0.33521905541419983,
|
| 1351 |
+
"learning_rate": 5.4878295598741325e-05,
|
| 1352 |
+
"loss": 0.0299,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 12.451612903225806,
|
| 1357 |
+
"grad_norm": 0.28460752964019775,
|
| 1358 |
+
"learning_rate": 5.4453428125361e-05,
|
| 1359 |
+
"loss": 0.0297,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 12.516129032258064,
|
| 1364 |
+
"grad_norm": 0.24392494559288025,
|
| 1365 |
+
"learning_rate": 5.40282362676094e-05,
|
| 1366 |
+
"loss": 0.0338,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 12.580645161290322,
|
| 1371 |
+
"grad_norm": 0.24771785736083984,
|
| 1372 |
+
"learning_rate": 5.360275099614319e-05,
|
| 1373 |
+
"loss": 0.0327,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 12.64516129032258,
|
| 1378 |
+
"grad_norm": 0.41986414790153503,
|
| 1379 |
+
"learning_rate": 5.317700330299107e-05,
|
| 1380 |
+
"loss": 0.0273,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 12.709677419354838,
|
| 1385 |
+
"grad_norm": 0.3345301151275635,
|
| 1386 |
+
"learning_rate": 5.275102419929632e-05,
|
| 1387 |
+
"loss": 0.0329,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 12.774193548387096,
|
| 1392 |
+
"grad_norm": 0.35422471165657043,
|
| 1393 |
+
"learning_rate": 5.232484471305804e-05,
|
| 1394 |
+
"loss": 0.035,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 12.838709677419354,
|
| 1399 |
+
"grad_norm": 0.2817692160606384,
|
| 1400 |
+
"learning_rate": 5.189849588687099e-05,
|
| 1401 |
+
"loss": 0.0288,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 12.903225806451612,
|
| 1406 |
+
"grad_norm": 0.4547789394855499,
|
| 1407 |
+
"learning_rate": 5.147200877566456e-05,
|
| 1408 |
+
"loss": 0.0347,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 12.967741935483872,
|
| 1413 |
+
"grad_norm": 0.2539784908294678,
|
| 1414 |
+
"learning_rate": 5.104541444444071e-05,
|
| 1415 |
+
"loss": 0.0352,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 13.03225806451613,
|
| 1420 |
+
"grad_norm": 0.2622417211532593,
|
| 1421 |
+
"learning_rate": 5.061874396601121e-05,
|
| 1422 |
+
"loss": 0.0296,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 13.096774193548388,
|
| 1427 |
+
"grad_norm": 0.3102482259273529,
|
| 1428 |
+
"learning_rate": 5.0192028418734335e-05,
|
| 1429 |
+
"loss": 0.0313,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 13.161290322580646,
|
| 1434 |
+
"grad_norm": 0.25369301438331604,
|
| 1435 |
+
"learning_rate": 4.9765298884251186e-05,
|
| 1436 |
+
"loss": 0.0309,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 13.225806451612904,
|
| 1441 |
+
"grad_norm": 0.3013876974582672,
|
| 1442 |
+
"learning_rate": 4.933858644522165e-05,
|
| 1443 |
+
"loss": 0.0372,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 13.290322580645162,
|
| 1448 |
+
"grad_norm": 0.28694114089012146,
|
| 1449 |
+
"learning_rate": 4.8911922183060356e-05,
|
| 1450 |
+
"loss": 0.0295,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 13.35483870967742,
|
| 1455 |
+
"grad_norm": 0.4296466112136841,
|
| 1456 |
+
"learning_rate": 4.848533717567283e-05,
|
| 1457 |
+
"loss": 0.0305,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 13.419354838709678,
|
| 1462 |
+
"grad_norm": 0.2966064512729645,
|
| 1463 |
+
"learning_rate": 4.805886249519172e-05,
|
| 1464 |
+
"loss": 0.0354,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 13.483870967741936,
|
| 1469 |
+
"grad_norm": 0.25592073798179626,
|
| 1470 |
+
"learning_rate": 4.763252920571351e-05,
|
| 1471 |
+
"loss": 0.0332,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 13.548387096774194,
|
| 1476 |
+
"grad_norm": 0.42708101868629456,
|
| 1477 |
+
"learning_rate": 4.720636836103593e-05,
|
| 1478 |
+
"loss": 0.0362,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 13.612903225806452,
|
| 1483 |
+
"grad_norm": 0.44891414046287537,
|
| 1484 |
+
"learning_rate": 4.6780411002395904e-05,
|
| 1485 |
+
"loss": 0.0312,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 13.67741935483871,
|
| 1490 |
+
"grad_norm": 0.33705928921699524,
|
| 1491 |
+
"learning_rate": 4.6354688156208616e-05,
|
| 1492 |
+
"loss": 0.0322,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 13.741935483870968,
|
| 1497 |
+
"grad_norm": 0.34952977299690247,
|
| 1498 |
+
"learning_rate": 4.592923083180752e-05,
|
| 1499 |
+
"loss": 0.0297,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 13.806451612903226,
|
| 1504 |
+
"grad_norm": 0.2684139013290405,
|
| 1505 |
+
"learning_rate": 4.55040700191857e-05,
|
| 1506 |
+
"loss": 0.0334,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 13.870967741935484,
|
| 1511 |
+
"grad_norm": 0.3220863342285156,
|
| 1512 |
+
"learning_rate": 4.507923668673847e-05,
|
| 1513 |
+
"loss": 0.0312,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 13.935483870967742,
|
| 1518 |
+
"grad_norm": 0.31924524903297424,
|
| 1519 |
+
"learning_rate": 4.465476177900782e-05,
|
| 1520 |
+
"loss": 0.0303,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 14.0,
|
| 1525 |
+
"grad_norm": 0.5949335694313049,
|
| 1526 |
+
"learning_rate": 4.423067621442831e-05,
|
| 1527 |
+
"loss": 0.0275,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 14.064516129032258,
|
| 1532 |
+
"grad_norm": 0.28924089670181274,
|
| 1533 |
+
"learning_rate": 4.380701088307501e-05,
|
| 1534 |
+
"loss": 0.0272,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 14.129032258064516,
|
| 1539 |
+
"grad_norm": 0.3126032054424286,
|
| 1540 |
+
"learning_rate": 4.338379664441357e-05,
|
| 1541 |
+
"loss": 0.0305,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 14.193548387096774,
|
| 1546 |
+
"grad_norm": 0.2669059634208679,
|
| 1547 |
+
"learning_rate": 4.2961064325052335e-05,
|
| 1548 |
+
"loss": 0.0286,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 14.258064516129032,
|
| 1553 |
+
"grad_norm": 0.3986091911792755,
|
| 1554 |
+
"learning_rate": 4.253884471649707e-05,
|
| 1555 |
+
"loss": 0.0305,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 14.32258064516129,
|
| 1560 |
+
"grad_norm": 0.397923082113266,
|
| 1561 |
+
"learning_rate": 4.2117168572907996e-05,
|
| 1562 |
+
"loss": 0.0324,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 14.387096774193548,
|
| 1567 |
+
"grad_norm": 0.3409479856491089,
|
| 1568 |
+
"learning_rate": 4.169606660885982e-05,
|
| 1569 |
+
"loss": 0.0276,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 14.451612903225806,
|
| 1574 |
+
"grad_norm": 0.2611437737941742,
|
| 1575 |
+
"learning_rate": 4.127556949710437e-05,
|
| 1576 |
+
"loss": 0.0283,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 14.516129032258064,
|
| 1581 |
+
"grad_norm": 0.3530644178390503,
|
| 1582 |
+
"learning_rate": 4.085570786633656e-05,
|
| 1583 |
+
"loss": 0.0278,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 14.580645161290322,
|
| 1588 |
+
"grad_norm": 0.23989705741405487,
|
| 1589 |
+
"learning_rate": 4.0436512298963305e-05,
|
| 1590 |
+
"loss": 0.0318,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 14.64516129032258,
|
| 1595 |
+
"grad_norm": 0.2740343511104584,
|
| 1596 |
+
"learning_rate": 4.001801332887595e-05,
|
| 1597 |
+
"loss": 0.0297,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 14.709677419354838,
|
| 1602 |
+
"grad_norm": 0.39286455512046814,
|
| 1603 |
+
"learning_rate": 3.960024143922625e-05,
|
| 1604 |
+
"loss": 0.0291,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 14.774193548387096,
|
| 1609 |
+
"grad_norm": 0.7122088074684143,
|
| 1610 |
+
"learning_rate": 3.918322706020593e-05,
|
| 1611 |
+
"loss": 0.0368,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 14.838709677419354,
|
| 1616 |
+
"grad_norm": 0.22896356880664825,
|
| 1617 |
+
"learning_rate": 3.876700056683026e-05,
|
| 1618 |
+
"loss": 0.027,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 14.903225806451612,
|
| 1623 |
+
"grad_norm": 0.3285376727581024,
|
| 1624 |
+
"learning_rate": 3.8351592276725425e-05,
|
| 1625 |
+
"loss": 0.0248,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 14.967741935483872,
|
| 1630 |
+
"grad_norm": 0.23626801371574402,
|
| 1631 |
+
"learning_rate": 3.793703244792038e-05,
|
| 1632 |
+
"loss": 0.0295,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 15.03225806451613,
|
| 1637 |
+
"grad_norm": 0.39834141731262207,
|
| 1638 |
+
"learning_rate": 3.7523351276642736e-05,
|
| 1639 |
+
"loss": 0.0292,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 15.096774193548388,
|
| 1644 |
+
"grad_norm": 0.29628074169158936,
|
| 1645 |
+
"learning_rate": 3.71105788951194e-05,
|
| 1646 |
+
"loss": 0.0275,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 15.161290322580646,
|
| 1651 |
+
"grad_norm": 0.18978668749332428,
|
| 1652 |
+
"learning_rate": 3.669874536938164e-05,
|
| 1653 |
+
"loss": 0.0265,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 15.225806451612904,
|
| 1658 |
+
"grad_norm": 0.30174311995506287,
|
| 1659 |
+
"learning_rate": 3.628788069707524e-05,
|
| 1660 |
+
"loss": 0.0266,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 15.290322580645162,
|
| 1665 |
+
"grad_norm": 0.24498514831066132,
|
| 1666 |
+
"learning_rate": 3.587801480527542e-05,
|
| 1667 |
+
"loss": 0.0259,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 15.35483870967742,
|
| 1672 |
+
"grad_norm": 0.3696959316730499,
|
| 1673 |
+
"learning_rate": 3.546917754830691e-05,
|
| 1674 |
+
"loss": 0.032,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 15.419354838709678,
|
| 1679 |
+
"grad_norm": 0.31626754999160767,
|
| 1680 |
+
"learning_rate": 3.5061398705569545e-05,
|
| 1681 |
+
"loss": 0.0264,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 15.483870967741936,
|
| 1686 |
+
"grad_norm": 0.5776691436767578,
|
| 1687 |
+
"learning_rate": 3.465470797936896e-05,
|
| 1688 |
+
"loss": 0.0244,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 15.548387096774194,
|
| 1693 |
+
"grad_norm": 0.25500577688217163,
|
| 1694 |
+
"learning_rate": 3.424913499275328e-05,
|
| 1695 |
+
"loss": 0.0304,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 15.612903225806452,
|
| 1700 |
+
"grad_norm": 0.3002663552761078,
|
| 1701 |
+
"learning_rate": 3.384470928735527e-05,
|
| 1702 |
+
"loss": 0.0287,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 15.67741935483871,
|
| 1707 |
+
"grad_norm": 0.45792222023010254,
|
| 1708 |
+
"learning_rate": 3.3441460321240605e-05,
|
| 1709 |
+
"loss": 0.0316,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 15.741935483870968,
|
| 1714 |
+
"grad_norm": 0.20462553203105927,
|
| 1715 |
+
"learning_rate": 3.303941746676212e-05,
|
| 1716 |
+
"loss": 0.0255,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 15.806451612903226,
|
| 1721 |
+
"grad_norm": 0.29612648487091064,
|
| 1722 |
+
"learning_rate": 3.263861000842043e-05,
|
| 1723 |
+
"loss": 0.0292,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 15.870967741935484,
|
| 1728 |
+
"grad_norm": 0.3627132475376129,
|
| 1729 |
+
"learning_rate": 3.223906714073083e-05,
|
| 1730 |
+
"loss": 0.0277,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 15.935483870967742,
|
| 1735 |
+
"grad_norm": 0.4241921901702881,
|
| 1736 |
+
"learning_rate": 3.1840817966096684e-05,
|
| 1737 |
+
"loss": 0.0323,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 16.0,
|
| 1742 |
+
"grad_norm": 0.4466724693775177,
|
| 1743 |
+
"learning_rate": 3.1443891492689834e-05,
|
| 1744 |
+
"loss": 0.0257,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 16.06451612903226,
|
| 1749 |
+
"grad_norm": 0.33882275223731995,
|
| 1750 |
+
"learning_rate": 3.104831663233746e-05,
|
| 1751 |
+
"loss": 0.0288,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 16.129032258064516,
|
| 1756 |
+
"grad_norm": 0.28266456723213196,
|
| 1757 |
+
"learning_rate": 3.065412219841635e-05,
|
| 1758 |
+
"loss": 0.0266,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 16.193548387096776,
|
| 1763 |
+
"grad_norm": 0.2942975163459778,
|
| 1764 |
+
"learning_rate": 3.026133690375398e-05,
|
| 1765 |
+
"loss": 0.026,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 16.258064516129032,
|
| 1770 |
+
"grad_norm": 0.28727394342422485,
|
| 1771 |
+
"learning_rate": 2.9869989358537253e-05,
|
| 1772 |
+
"loss": 0.0282,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 16.322580645161292,
|
| 1777 |
+
"grad_norm": 0.28704243898391724,
|
| 1778 |
+
"learning_rate": 2.9480108068228397e-05,
|
| 1779 |
+
"loss": 0.0285,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 16.387096774193548,
|
| 1784 |
+
"grad_norm": 0.3424932360649109,
|
| 1785 |
+
"learning_rate": 2.9091721431488784e-05,
|
| 1786 |
+
"loss": 0.0298,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 16.451612903225808,
|
| 1791 |
+
"grad_norm": 0.4286734461784363,
|
| 1792 |
+
"learning_rate": 2.8704857738110368e-05,
|
| 1793 |
+
"loss": 0.0244,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 16.516129032258064,
|
| 1798 |
+
"grad_norm": 0.2728700339794159,
|
| 1799 |
+
"learning_rate": 2.8319545166954963e-05,
|
| 1800 |
+
"loss": 0.0265,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 16.580645161290324,
|
| 1805 |
+
"grad_norm": 0.5243811011314392,
|
| 1806 |
+
"learning_rate": 2.793581178390188e-05,
|
| 1807 |
+
"loss": 0.0242,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 16.64516129032258,
|
| 1812 |
+
"grad_norm": 0.3646429479122162,
|
| 1813 |
+
"learning_rate": 2.7553685539803488e-05,
|
| 1814 |
+
"loss": 0.027,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 16.70967741935484,
|
| 1819 |
+
"grad_norm": 0.24419990181922913,
|
| 1820 |
+
"learning_rate": 2.717319426844942e-05,
|
| 1821 |
+
"loss": 0.0274,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 16.774193548387096,
|
| 1826 |
+
"grad_norm": 0.3144078254699707,
|
| 1827 |
+
"learning_rate": 2.679436568453904e-05,
|
| 1828 |
+
"loss": 0.0266,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 16.838709677419356,
|
| 1833 |
+
"grad_norm": 0.3204503655433655,
|
| 1834 |
+
"learning_rate": 2.64172273816629e-05,
|
| 1835 |
+
"loss": 0.027,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 16.903225806451612,
|
| 1840 |
+
"grad_norm": 0.2552044093608856,
|
| 1841 |
+
"learning_rate": 2.6041806830292647e-05,
|
| 1842 |
+
"loss": 0.0269,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 16.967741935483872,
|
| 1847 |
+
"grad_norm": 0.25380566716194153,
|
| 1848 |
+
"learning_rate": 2.5668131375780276e-05,
|
| 1849 |
+
"loss": 0.0231,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 17.032258064516128,
|
| 1854 |
+
"grad_norm": 0.3057281970977783,
|
| 1855 |
+
"learning_rate": 2.5296228236366194e-05,
|
| 1856 |
+
"loss": 0.0266,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 17.096774193548388,
|
| 1861 |
+
"grad_norm": 0.26944369077682495,
|
| 1862 |
+
"learning_rate": 2.492612450119669e-05,
|
| 1863 |
+
"loss": 0.0264,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 17.161290322580644,
|
| 1868 |
+
"grad_norm": 0.31917497515678406,
|
| 1869 |
+
"learning_rate": 2.455784712835084e-05,
|
| 1870 |
+
"loss": 0.0265,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 17.225806451612904,
|
| 1875 |
+
"grad_norm": 0.2606102228164673,
|
| 1876 |
+
"learning_rate": 2.4191422942876803e-05,
|
| 1877 |
+
"loss": 0.0238,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 17.29032258064516,
|
| 1882 |
+
"grad_norm": 0.25582873821258545,
|
| 1883 |
+
"learning_rate": 2.3826878634838e-05,
|
| 1884 |
+
"loss": 0.024,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 17.35483870967742,
|
| 1889 |
+
"grad_norm": 0.25627753138542175,
|
| 1890 |
+
"learning_rate": 2.3464240757368944e-05,
|
| 1891 |
+
"loss": 0.0261,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 17.419354838709676,
|
| 1896 |
+
"grad_norm": 0.2993682622909546,
|
| 1897 |
+
"learning_rate": 2.3103535724741233e-05,
|
| 1898 |
+
"loss": 0.0267,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 17.483870967741936,
|
| 1903 |
+
"grad_norm": 0.24699701368808746,
|
| 1904 |
+
"learning_rate": 2.2744789810439422e-05,
|
| 1905 |
+
"loss": 0.0246,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 17.548387096774192,
|
| 1910 |
+
"grad_norm": 0.24654793739318848,
|
| 1911 |
+
"learning_rate": 2.2388029145247396e-05,
|
| 1912 |
+
"loss": 0.0216,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 17.612903225806452,
|
| 1917 |
+
"grad_norm": 0.3724614083766937,
|
| 1918 |
+
"learning_rate": 2.2033279715345006e-05,
|
| 1919 |
+
"loss": 0.0264,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 17.677419354838708,
|
| 1924 |
+
"grad_norm": 0.2768930196762085,
|
| 1925 |
+
"learning_rate": 2.1680567360415155e-05,
|
| 1926 |
+
"loss": 0.0264,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 17.741935483870968,
|
| 1931 |
+
"grad_norm": 0.1919165402650833,
|
| 1932 |
+
"learning_rate": 2.1329917771761805e-05,
|
| 1933 |
+
"loss": 0.0218,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 17.806451612903224,
|
| 1938 |
+
"grad_norm": 0.15743491053581238,
|
| 1939 |
+
"learning_rate": 2.0981356490438502e-05,
|
| 1940 |
+
"loss": 0.0213,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 17.870967741935484,
|
| 1945 |
+
"grad_norm": 0.3243708312511444,
|
| 1946 |
+
"learning_rate": 2.0634908905388123e-05,
|
| 1947 |
+
"loss": 0.0221,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 17.93548387096774,
|
| 1952 |
+
"grad_norm": 1.23497474193573,
|
| 1953 |
+
"learning_rate": 2.02906002515934e-05,
|
| 1954 |
+
"loss": 0.0252,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 18.0,
|
| 1959 |
+
"grad_norm": 0.3177509903907776,
|
| 1960 |
+
"learning_rate": 1.9948455608238985e-05,
|
| 1961 |
+
"loss": 0.0272,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 18.06451612903226,
|
| 1966 |
+
"grad_norm": 0.29627230763435364,
|
| 1967 |
+
"learning_rate": 1.960849989688463e-05,
|
| 1968 |
+
"loss": 0.0238,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 18.129032258064516,
|
| 1973 |
+
"grad_norm": 0.47126612067222595,
|
| 1974 |
+
"learning_rate": 1.927075787964985e-05,
|
| 1975 |
+
"loss": 0.0226,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 18.193548387096776,
|
| 1980 |
+
"grad_norm": 0.22074201703071594,
|
| 1981 |
+
"learning_rate": 1.8935254157410403e-05,
|
| 1982 |
+
"loss": 0.0227,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 18.258064516129032,
|
| 1987 |
+
"grad_norm": 0.23865483701229095,
|
| 1988 |
+
"learning_rate": 1.8602013168006293e-05,
|
| 1989 |
+
"loss": 0.0232,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 18.322580645161292,
|
| 1994 |
+
"grad_norm": 0.24160803854465485,
|
| 1995 |
+
"learning_rate": 1.827105918446178e-05,
|
| 1996 |
+
"loss": 0.0226,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 18.387096774193548,
|
| 2001 |
+
"grad_norm": 0.29074233770370483,
|
| 2002 |
+
"learning_rate": 1.794241631321727e-05,
|
| 2003 |
+
"loss": 0.0215,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 18.451612903225808,
|
| 2008 |
+
"grad_norm": 0.2850009799003601,
|
| 2009 |
+
"learning_rate": 1.7616108492373545e-05,
|
| 2010 |
+
"loss": 0.0209,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 18.516129032258064,
|
| 2015 |
+
"grad_norm": 0.39563989639282227,
|
| 2016 |
+
"learning_rate": 1.7292159489948012e-05,
|
| 2017 |
+
"loss": 0.0234,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 18.580645161290324,
|
| 2022 |
+
"grad_norm": 0.2690832316875458,
|
| 2023 |
+
"learning_rate": 1.697059290214356e-05,
|
| 2024 |
+
"loss": 0.0251,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 18.64516129032258,
|
| 2029 |
+
"grad_norm": 0.21451742947101593,
|
| 2030 |
+
"learning_rate": 1.6651432151629725e-05,
|
| 2031 |
+
"loss": 0.0251,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 18.70967741935484,
|
| 2036 |
+
"grad_norm": 0.5447711944580078,
|
| 2037 |
+
"learning_rate": 1.6334700485836714e-05,
|
| 2038 |
+
"loss": 0.0241,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 18.774193548387096,
|
| 2043 |
+
"grad_norm": 0.26067161560058594,
|
| 2044 |
+
"learning_rate": 1.6020420975261956e-05,
|
| 2045 |
+
"loss": 0.0279,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 18.838709677419356,
|
| 2050 |
+
"grad_norm": 0.3259269595146179,
|
| 2051 |
+
"learning_rate": 1.570861651178978e-05,
|
| 2052 |
+
"loss": 0.0219,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 18.903225806451612,
|
| 2057 |
+
"grad_norm": 0.2738400399684906,
|
| 2058 |
+
"learning_rate": 1.5399309807023944e-05,
|
| 2059 |
+
"loss": 0.0205,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 18.967741935483872,
|
| 2064 |
+
"grad_norm": 0.27828454971313477,
|
| 2065 |
+
"learning_rate": 1.5092523390633267e-05,
|
| 2066 |
+
"loss": 0.0226,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 19.032258064516128,
|
| 2071 |
+
"grad_norm": 0.23453515768051147,
|
| 2072 |
+
"learning_rate": 1.4788279608710732e-05,
|
| 2073 |
+
"loss": 0.025,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 19.096774193548388,
|
| 2078 |
+
"grad_norm": 0.3187068998813629,
|
| 2079 |
+
"learning_rate": 1.4486600622145652e-05,
|
| 2080 |
+
"loss": 0.0242,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 19.161290322580644,
|
| 2085 |
+
"grad_norm": 0.29299116134643555,
|
| 2086 |
+
"learning_rate": 1.4187508405009637e-05,
|
| 2087 |
+
"loss": 0.0216,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 19.225806451612904,
|
| 2092 |
+
"grad_norm": 0.25096479058265686,
|
| 2093 |
+
"learning_rate": 1.3891024742955877e-05,
|
| 2094 |
+
"loss": 0.0217,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 19.29032258064516,
|
| 2099 |
+
"grad_norm": 0.34088948369026184,
|
| 2100 |
+
"learning_rate": 1.3597171231632433e-05,
|
| 2101 |
+
"loss": 0.0233,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 19.35483870967742,
|
| 2106 |
+
"grad_norm": 0.255126416683197,
|
| 2107 |
+
"learning_rate": 1.3305969275109098e-05,
|
| 2108 |
+
"loss": 0.023,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 19.419354838709676,
|
| 2113 |
+
"grad_norm": 0.2508387863636017,
|
| 2114 |
+
"learning_rate": 1.3017440084318421e-05,
|
| 2115 |
+
"loss": 0.0205,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 19.483870967741936,
|
| 2120 |
+
"grad_norm": 0.21891236305236816,
|
| 2121 |
+
"learning_rate": 1.2731604675510728e-05,
|
| 2122 |
+
"loss": 0.0221,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 19.548387096774192,
|
| 2127 |
+
"grad_norm": 0.3564584255218506,
|
| 2128 |
+
"learning_rate": 1.2448483868723216e-05,
|
| 2129 |
+
"loss": 0.0243,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 19.612903225806452,
|
| 2134 |
+
"grad_norm": 0.2741142809391022,
|
| 2135 |
+
"learning_rate": 1.2168098286263557e-05,
|
| 2136 |
+
"loss": 0.022,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 19.677419354838708,
|
| 2141 |
+
"grad_norm": 0.40683725476264954,
|
| 2142 |
+
"learning_rate": 1.189046835120769e-05,
|
| 2143 |
+
"loss": 0.0265,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 19.741935483870968,
|
| 2148 |
+
"grad_norm": 0.3363529145717621,
|
| 2149 |
+
"learning_rate": 1.1615614285912301e-05,
|
| 2150 |
+
"loss": 0.0255,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 19.806451612903224,
|
| 2155 |
+
"grad_norm": 0.17003341019153595,
|
| 2156 |
+
"learning_rate": 1.134355611054176e-05,
|
| 2157 |
+
"loss": 0.0214,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 19.870967741935484,
|
| 2162 |
+
"grad_norm": 0.27854353189468384,
|
| 2163 |
+
"learning_rate": 1.1074313641609935e-05,
|
| 2164 |
+
"loss": 0.023,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 19.93548387096774,
|
| 2169 |
+
"grad_norm": 0.3035428524017334,
|
| 2170 |
+
"learning_rate": 1.0807906490536751e-05,
|
| 2171 |
+
"loss": 0.0233,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 20.0,
|
| 2176 |
+
"grad_norm": 0.3595190644264221,
|
| 2177 |
+
"learning_rate": 1.0544354062219658e-05,
|
| 2178 |
+
"loss": 0.0228,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 20.06451612903226,
|
| 2183 |
+
"grad_norm": 0.2843579649925232,
|
| 2184 |
+
"learning_rate": 1.0283675553620281e-05,
|
| 2185 |
+
"loss": 0.0192,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 20.129032258064516,
|
| 2190 |
+
"grad_norm": 0.4682447910308838,
|
| 2191 |
+
"learning_rate": 1.0025889952366085e-05,
|
| 2192 |
+
"loss": 0.0243,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 20.193548387096776,
|
| 2197 |
+
"grad_norm": 0.1721201092004776,
|
| 2198 |
+
"learning_rate": 9.77101603536728e-06,
|
| 2199 |
+
"loss": 0.0234,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 20.258064516129032,
|
| 2204 |
+
"grad_norm": 0.2416217029094696,
|
| 2205 |
+
"learning_rate": 9.519072367449239e-06,
|
| 2206 |
+
"loss": 0.0189,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 20.322580645161292,
|
| 2211 |
+
"grad_norm": 0.28073132038116455,
|
| 2212 |
+
"learning_rate": 9.270077300000145e-06,
|
| 2213 |
+
"loss": 0.021,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 20.387096774193548,
|
| 2218 |
+
"grad_norm": 0.2682770788669586,
|
| 2219 |
+
"learning_rate": 9.02404896963433e-06,
|
| 2220 |
+
"loss": 0.0201,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 20.451612903225808,
|
| 2225 |
+
"grad_norm": 0.29497283697128296,
|
| 2226 |
+
"learning_rate": 8.781005296871248e-06,
|
| 2227 |
+
"loss": 0.0305,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 20.516129032258064,
|
| 2232 |
+
"grad_norm": 0.2409420758485794,
|
| 2233 |
+
"learning_rate": 8.540963984830075e-06,
|
| 2234 |
+
"loss": 0.0206,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 20.580645161290324,
|
| 2239 |
+
"grad_norm": 0.23492980003356934,
|
| 2240 |
+
"learning_rate": 8.303942517940327e-06,
|
| 2241 |
+
"loss": 0.023,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 20.64516129032258,
|
| 2246 |
+
"grad_norm": 0.18785379827022552,
|
| 2247 |
+
"learning_rate": 8.069958160668256e-06,
|
| 2248 |
+
"loss": 0.0217,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 20.70967741935484,
|
| 2253 |
+
"grad_norm": 0.2183133363723755,
|
| 2254 |
+
"learning_rate": 7.839027956259282e-06,
|
| 2255 |
+
"loss": 0.0213,
|
| 2256 |
+
"step": 3210
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 20.774193548387096,
|
| 2260 |
+
"grad_norm": 0.20956051349639893,
|
| 2261 |
+
"learning_rate": 7.6111687254966705e-06,
|
| 2262 |
+
"loss": 0.0188,
|
| 2263 |
+
"step": 3220
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 20.838709677419356,
|
| 2267 |
+
"grad_norm": 0.2115253210067749,
|
| 2268 |
+
"learning_rate": 7.386397065476209e-06,
|
| 2269 |
+
"loss": 0.0219,
|
| 2270 |
+
"step": 3230
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 20.903225806451612,
|
| 2274 |
+
"grad_norm": 0.27826544642448425,
|
| 2275 |
+
"learning_rate": 7.164729348397409e-06,
|
| 2276 |
+
"loss": 0.021,
|
| 2277 |
+
"step": 3240
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 20.967741935483872,
|
| 2281 |
+
"grad_norm": 0.8272599577903748,
|
| 2282 |
+
"learning_rate": 6.946181720370825e-06,
|
| 2283 |
+
"loss": 0.0197,
|
| 2284 |
+
"step": 3250
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 21.032258064516128,
|
| 2288 |
+
"grad_norm": 0.31838661432266235,
|
| 2289 |
+
"learning_rate": 6.73077010024214e-06,
|
| 2290 |
+
"loss": 0.025,
|
| 2291 |
+
"step": 3260
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 21.096774193548388,
|
| 2295 |
+
"grad_norm": 0.32917433977127075,
|
| 2296 |
+
"learning_rate": 6.518510178432491e-06,
|
| 2297 |
+
"loss": 0.0209,
|
| 2298 |
+
"step": 3270
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 21.161290322580644,
|
| 2302 |
+
"grad_norm": 0.21181564033031464,
|
| 2303 |
+
"learning_rate": 6.309417415795732e-06,
|
| 2304 |
+
"loss": 0.0203,
|
| 2305 |
+
"step": 3280
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 21.225806451612904,
|
| 2309 |
+
"grad_norm": 0.13702601194381714,
|
| 2310 |
+
"learning_rate": 6.1035070424921905e-06,
|
| 2311 |
+
"loss": 0.0238,
|
| 2312 |
+
"step": 3290
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 21.29032258064516,
|
| 2316 |
+
"grad_norm": 0.5193454027175903,
|
| 2317 |
+
"learning_rate": 5.900794056879317e-06,
|
| 2318 |
+
"loss": 0.021,
|
| 2319 |
+
"step": 3300
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 21.35483870967742,
|
| 2323 |
+
"grad_norm": 0.2874934673309326,
|
| 2324 |
+
"learning_rate": 5.7012932244192745e-06,
|
| 2325 |
+
"loss": 0.0202,
|
| 2326 |
+
"step": 3310
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 21.419354838709676,
|
| 2330 |
+
"grad_norm": 0.20375904440879822,
|
| 2331 |
+
"learning_rate": 5.505019076603352e-06,
|
| 2332 |
+
"loss": 0.0206,
|
| 2333 |
+
"step": 3320
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 21.483870967741936,
|
| 2337 |
+
"grad_norm": 0.25239813327789307,
|
| 2338 |
+
"learning_rate": 5.311985909893591e-06,
|
| 2339 |
+
"loss": 0.0208,
|
| 2340 |
+
"step": 3330
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 21.548387096774192,
|
| 2344 |
+
"grad_norm": 0.1695331186056137,
|
| 2345 |
+
"learning_rate": 5.122207784681337e-06,
|
| 2346 |
+
"loss": 0.0241,
|
| 2347 |
+
"step": 3340
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 21.612903225806452,
|
| 2351 |
+
"grad_norm": 0.2161603420972824,
|
| 2352 |
+
"learning_rate": 4.935698524263216e-06,
|
| 2353 |
+
"loss": 0.0214,
|
| 2354 |
+
"step": 3350
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 21.677419354838708,
|
| 2358 |
+
"grad_norm": 0.3184122145175934,
|
| 2359 |
+
"learning_rate": 4.7524717138341335e-06,
|
| 2360 |
+
"loss": 0.0205,
|
| 2361 |
+
"step": 3360
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 21.741935483870968,
|
| 2365 |
+
"grad_norm": 0.24667605757713318,
|
| 2366 |
+
"learning_rate": 4.5725406994978235e-06,
|
| 2367 |
+
"loss": 0.0182,
|
| 2368 |
+
"step": 3370
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 21.806451612903224,
|
| 2372 |
+
"grad_norm": 0.21050481498241425,
|
| 2373 |
+
"learning_rate": 4.395918587294701e-06,
|
| 2374 |
+
"loss": 0.0223,
|
| 2375 |
+
"step": 3380
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 21.870967741935484,
|
| 2379 |
+
"grad_norm": 0.31748640537261963,
|
| 2380 |
+
"learning_rate": 4.222618242247195e-06,
|
| 2381 |
+
"loss": 0.0164,
|
| 2382 |
+
"step": 3390
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 21.93548387096774,
|
| 2386 |
+
"grad_norm": 0.23009532690048218,
|
| 2387 |
+
"learning_rate": 4.052652287422754e-06,
|
| 2388 |
+
"loss": 0.0223,
|
| 2389 |
+
"step": 3400
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 22.0,
|
| 2393 |
+
"grad_norm": 0.46855324506759644,
|
| 2394 |
+
"learning_rate": 3.886033103014286e-06,
|
| 2395 |
+
"loss": 0.0228,
|
| 2396 |
+
"step": 3410
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 22.06451612903226,
|
| 2400 |
+
"grad_norm": 0.3878401219844818,
|
| 2401 |
+
"learning_rate": 3.7227728254384752e-06,
|
| 2402 |
+
"loss": 0.0214,
|
| 2403 |
+
"step": 3420
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 22.129032258064516,
|
| 2407 |
+
"grad_norm": 0.26421085000038147,
|
| 2408 |
+
"learning_rate": 3.5628833464517398e-06,
|
| 2409 |
+
"loss": 0.0214,
|
| 2410 |
+
"step": 3430
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 22.193548387096776,
|
| 2414 |
+
"grad_norm": 0.2137315571308136,
|
| 2415 |
+
"learning_rate": 3.4063763122840654e-06,
|
| 2416 |
+
"loss": 0.0185,
|
| 2417 |
+
"step": 3440
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 22.258064516129032,
|
| 2421 |
+
"grad_norm": 0.20358504354953766,
|
| 2422 |
+
"learning_rate": 3.2532631227906562e-06,
|
| 2423 |
+
"loss": 0.0205,
|
| 2424 |
+
"step": 3450
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 22.322580645161292,
|
| 2428 |
+
"grad_norm": 0.2786470055580139,
|
| 2429 |
+
"learning_rate": 3.103554930621644e-06,
|
| 2430 |
+
"loss": 0.0225,
|
| 2431 |
+
"step": 3460
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 22.387096774193548,
|
| 2435 |
+
"grad_norm": 0.25646957755088806,
|
| 2436 |
+
"learning_rate": 2.9572626404096917e-06,
|
| 2437 |
+
"loss": 0.0191,
|
| 2438 |
+
"step": 3470
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 22.451612903225808,
|
| 2442 |
+
"grad_norm": 0.15664377808570862,
|
| 2443 |
+
"learning_rate": 2.8143969079757083e-06,
|
| 2444 |
+
"loss": 0.02,
|
| 2445 |
+
"step": 3480
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 22.516129032258064,
|
| 2449 |
+
"grad_norm": 0.21855230629444122,
|
| 2450 |
+
"learning_rate": 2.6749681395527194e-06,
|
| 2451 |
+
"loss": 0.0187,
|
| 2452 |
+
"step": 3490
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 22.580645161290324,
|
| 2456 |
+
"grad_norm": 0.15705305337905884,
|
| 2457 |
+
"learning_rate": 2.538986491027845e-06,
|
| 2458 |
+
"loss": 0.0198,
|
| 2459 |
+
"step": 3500
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 22.64516129032258,
|
| 2463 |
+
"grad_norm": 0.3838616609573364,
|
| 2464 |
+
"learning_rate": 2.4064618672025917e-06,
|
| 2465 |
+
"loss": 0.0203,
|
| 2466 |
+
"step": 3510
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 22.70967741935484,
|
| 2470 |
+
"grad_norm": 0.1712443083524704,
|
| 2471 |
+
"learning_rate": 2.2774039210713584e-06,
|
| 2472 |
+
"loss": 0.0205,
|
| 2473 |
+
"step": 3520
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 22.774193548387096,
|
| 2477 |
+
"grad_norm": 0.3191426396369934,
|
| 2478 |
+
"learning_rate": 2.151822053118352e-06,
|
| 2479 |
+
"loss": 0.0204,
|
| 2480 |
+
"step": 3530
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 22.838709677419356,
|
| 2484 |
+
"grad_norm": 0.21879324316978455,
|
| 2485 |
+
"learning_rate": 2.029725410632821e-06,
|
| 2486 |
+
"loss": 0.019,
|
| 2487 |
+
"step": 3540
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 22.903225806451612,
|
| 2491 |
+
"grad_norm": 0.25989222526550293,
|
| 2492 |
+
"learning_rate": 1.9111228870428267e-06,
|
| 2493 |
+
"loss": 0.019,
|
| 2494 |
+
"step": 3550
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 22.967741935483872,
|
| 2498 |
+
"grad_norm": 0.22728225588798523,
|
| 2499 |
+
"learning_rate": 1.7960231212674095e-06,
|
| 2500 |
+
"loss": 0.019,
|
| 2501 |
+
"step": 3560
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 23.032258064516128,
|
| 2505 |
+
"grad_norm": 0.3381948173046112,
|
| 2506 |
+
"learning_rate": 1.684434497087356e-06,
|
| 2507 |
+
"loss": 0.0204,
|
| 2508 |
+
"step": 3570
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 23.096774193548388,
|
| 2512 |
+
"grad_norm": 0.3168867230415344,
|
| 2513 |
+
"learning_rate": 1.5763651425345306e-06,
|
| 2514 |
+
"loss": 0.0173,
|
| 2515 |
+
"step": 3580
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 23.161290322580644,
|
| 2519 |
+
"grad_norm": 0.4736108183860779,
|
| 2520 |
+
"learning_rate": 1.4718229292998265e-06,
|
| 2521 |
+
"loss": 0.0193,
|
| 2522 |
+
"step": 3590
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 23.225806451612904,
|
| 2526 |
+
"grad_norm": 0.1927083134651184,
|
| 2527 |
+
"learning_rate": 1.3708154721598142e-06,
|
| 2528 |
+
"loss": 0.0162,
|
| 2529 |
+
"step": 3600
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 23.29032258064516,
|
| 2533 |
+
"grad_norm": 0.26042094826698303,
|
| 2534 |
+
"learning_rate": 1.2733501284220573e-06,
|
| 2535 |
+
"loss": 0.019,
|
| 2536 |
+
"step": 3610
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 23.35483870967742,
|
| 2540 |
+
"grad_norm": 0.3572435975074768,
|
| 2541 |
+
"learning_rate": 1.1794339973892509e-06,
|
| 2542 |
+
"loss": 0.0187,
|
| 2543 |
+
"step": 3620
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 23.419354838709676,
|
| 2547 |
+
"grad_norm": 0.31472429633140564,
|
| 2548 |
+
"learning_rate": 1.0890739198420763e-06,
|
| 2549 |
+
"loss": 0.0223,
|
| 2550 |
+
"step": 3630
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 23.483870967741936,
|
| 2554 |
+
"grad_norm": 0.28377604484558105,
|
| 2555 |
+
"learning_rate": 1.0022764775409532e-06,
|
| 2556 |
+
"loss": 0.0187,
|
| 2557 |
+
"step": 3640
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 23.548387096774192,
|
| 2561 |
+
"grad_norm": 0.2872464656829834,
|
| 2562 |
+
"learning_rate": 9.190479927466023e-07,
|
| 2563 |
+
"loss": 0.0195,
|
| 2564 |
+
"step": 3650
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 23.612903225806452,
|
| 2568 |
+
"grad_norm": 0.2037927806377411,
|
| 2569 |
+
"learning_rate": 8.393945277595682e-07,
|
| 2570 |
+
"loss": 0.0209,
|
| 2571 |
+
"step": 3660
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 23.677419354838708,
|
| 2575 |
+
"grad_norm": 0.21536409854888916,
|
| 2576 |
+
"learning_rate": 7.633218844786238e-07,
|
| 2577 |
+
"loss": 0.0209,
|
| 2578 |
+
"step": 3670
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 23.741935483870968,
|
| 2582 |
+
"grad_norm": 0.19614161550998688,
|
| 2583 |
+
"learning_rate": 6.908356039781516e-07,
|
| 2584 |
+
"loss": 0.0188,
|
| 2585 |
+
"step": 3680
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 23.806451612903224,
|
| 2589 |
+
"grad_norm": 0.2418230175971985,
|
| 2590 |
+
"learning_rate": 6.219409661045839e-07,
|
| 2591 |
+
"loss": 0.0209,
|
| 2592 |
+
"step": 3690
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 23.870967741935484,
|
| 2596 |
+
"grad_norm": 0.41142016649246216,
|
| 2597 |
+
"learning_rate": 5.566429890917601e-07,
|
| 2598 |
+
"loss": 0.0201,
|
| 2599 |
+
"step": 3700
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 23.93548387096774,
|
| 2603 |
+
"grad_norm": 0.22285668551921844,
|
| 2604 |
+
"learning_rate": 4.949464291954642e-07,
|
| 2605 |
+
"loss": 0.0183,
|
| 2606 |
+
"step": 3710
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 24.0,
|
| 2610 |
+
"grad_norm": 0.27075517177581787,
|
| 2611 |
+
"learning_rate": 4.368557803469342e-07,
|
| 2612 |
+
"loss": 0.0185,
|
| 2613 |
+
"step": 3720
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 24.06451612903226,
|
| 2617 |
+
"grad_norm": 0.31757959723472595,
|
| 2618 |
+
"learning_rate": 3.823752738255471e-07,
|
| 2619 |
+
"loss": 0.022,
|
| 2620 |
+
"step": 3730
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 24.129032258064516,
|
| 2624 |
+
"grad_norm": 0.18830479681491852,
|
| 2625 |
+
"learning_rate": 3.315088779506259e-07,
|
| 2626 |
+
"loss": 0.0198,
|
| 2627 |
+
"step": 3740
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 24.193548387096776,
|
| 2631 |
+
"grad_norm": 0.5040279626846313,
|
| 2632 |
+
"learning_rate": 2.8426029779236586e-07,
|
| 2633 |
+
"loss": 0.0241,
|
| 2634 |
+
"step": 3750
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 24.258064516129032,
|
| 2638 |
+
"grad_norm": 0.7920020222663879,
|
| 2639 |
+
"learning_rate": 2.406329749019942e-07,
|
| 2640 |
+
"loss": 0.0213,
|
| 2641 |
+
"step": 3760
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 24.322580645161292,
|
| 2645 |
+
"grad_norm": 0.18160688877105713,
|
| 2646 |
+
"learning_rate": 2.0063008706103782e-07,
|
| 2647 |
+
"loss": 0.0186,
|
| 2648 |
+
"step": 3770
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 24.387096774193548,
|
| 2652 |
+
"grad_norm": 0.29395878314971924,
|
| 2653 |
+
"learning_rate": 1.6425454804991913e-07,
|
| 2654 |
+
"loss": 0.0198,
|
| 2655 |
+
"step": 3780
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 24.451612903225808,
|
| 2659 |
+
"grad_norm": 0.2948833107948303,
|
| 2660 |
+
"learning_rate": 1.3150900743566509e-07,
|
| 2661 |
+
"loss": 0.02,
|
| 2662 |
+
"step": 3790
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 24.516129032258064,
|
| 2666 |
+
"grad_norm": 0.40248140692710876,
|
| 2667 |
+
"learning_rate": 1.0239585037894462e-07,
|
| 2668 |
+
"loss": 0.0179,
|
| 2669 |
+
"step": 3800
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 24.580645161290324,
|
| 2673 |
+
"grad_norm": 0.2159591168165207,
|
| 2674 |
+
"learning_rate": 7.691719746033e-08,
|
| 2675 |
+
"loss": 0.0151,
|
| 2676 |
+
"step": 3810
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 24.64516129032258,
|
| 2680 |
+
"grad_norm": 0.3262782692909241,
|
| 2681 |
+
"learning_rate": 5.507490452582587e-08,
|
| 2682 |
+
"loss": 0.0214,
|
| 2683 |
+
"step": 3820
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 24.70967741935484,
|
| 2687 |
+
"grad_norm": 0.47080403566360474,
|
| 2688 |
+
"learning_rate": 3.687056255169963e-08,
|
| 2689 |
+
"loss": 0.0234,
|
| 2690 |
+
"step": 3830
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 24.774193548387096,
|
| 2694 |
+
"grad_norm": 0.5033960342407227,
|
| 2695 |
+
"learning_rate": 2.2305497528601892e-08,
|
| 2696 |
+
"loss": 0.023,
|
| 2697 |
+
"step": 3840
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 24.838709677419356,
|
| 2701 |
+
"grad_norm": 0.3430034816265106,
|
| 2702 |
+
"learning_rate": 1.1380770364960436e-08,
|
| 2703 |
+
"loss": 0.0234,
|
| 2704 |
+
"step": 3850
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 24.903225806451612,
|
| 2708 |
+
"grad_norm": 0.23608365654945374,
|
| 2709 |
+
"learning_rate": 4.0971768097253224e-09,
|
| 2710 |
+
"loss": 0.0199,
|
| 2711 |
+
"step": 3860
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 24.967741935483872,
|
| 2715 |
+
"grad_norm": 0.6302427649497986,
|
| 2716 |
+
"learning_rate": 4.55247394404168e-10,
|
| 2717 |
+
"loss": 0.0207,
|
| 2718 |
+
"step": 3870
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 25.0,
|
| 2722 |
+
"step": 3875,
|
| 2723 |
+
"total_flos": 4.172300356853376e+17,
|
| 2724 |
+
"train_loss": 0.051802249112436846,
|
| 2725 |
+
"train_runtime": 4196.5277,
|
| 2726 |
+
"train_samples_per_second": 45.103,
|
| 2727 |
+
"train_steps_per_second": 0.923
|
| 2728 |
+
}
|
| 2729 |
+
],
|
| 2730 |
+
"logging_steps": 10,
|
| 2731 |
+
"max_steps": 3875,
|
| 2732 |
+
"num_input_tokens_seen": 0,
|
| 2733 |
+
"num_train_epochs": 25,
|
| 2734 |
+
"save_steps": 10000,
|
| 2735 |
+
"stateful_callbacks": {
|
| 2736 |
+
"TrainerControl": {
|
| 2737 |
+
"args": {
|
| 2738 |
+
"should_epoch_stop": false,
|
| 2739 |
+
"should_evaluate": false,
|
| 2740 |
+
"should_log": false,
|
| 2741 |
+
"should_save": true,
|
| 2742 |
+
"should_training_stop": true
|
| 2743 |
+
},
|
| 2744 |
+
"attributes": {}
|
| 2745 |
+
}
|
| 2746 |
+
},
|
| 2747 |
+
"total_flos": 4.172300356853376e+17,
|
| 2748 |
+
"train_batch_size": 49,
|
| 2749 |
+
"trial_name": null,
|
| 2750 |
+
"trial_params": null
|
| 2751 |
+
}
|