Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2366 -0
trainer_state.json
ADDED
@@ -0,0 +1,2366 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 19.92814371257485,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 3328,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.059880239520958084,
|
13 |
+
"grad_norm": 4.727615833282471,
|
14 |
+
"learning_rate": 1.1976047904191617e-05,
|
15 |
+
"loss": 1.062,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.11976047904191617,
|
20 |
+
"grad_norm": 2.2185091972351074,
|
21 |
+
"learning_rate": 2.3952095808383234e-05,
|
22 |
+
"loss": 0.6786,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.17964071856287425,
|
27 |
+
"grad_norm": 1.7475155591964722,
|
28 |
+
"learning_rate": 3.592814371257485e-05,
|
29 |
+
"loss": 0.3714,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.23952095808383234,
|
34 |
+
"grad_norm": 0.7705522179603577,
|
35 |
+
"learning_rate": 4.790419161676647e-05,
|
36 |
+
"loss": 0.2662,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.2994011976047904,
|
41 |
+
"grad_norm": 0.5850515365600586,
|
42 |
+
"learning_rate": 5.988023952095808e-05,
|
43 |
+
"loss": 0.2129,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.3592814371257485,
|
48 |
+
"grad_norm": 0.5091089010238647,
|
49 |
+
"learning_rate": 7.18562874251497e-05,
|
50 |
+
"loss": 0.1831,
|
51 |
+
"step": 60
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.41916167664670656,
|
55 |
+
"grad_norm": 0.8252758383750916,
|
56 |
+
"learning_rate": 8.383233532934131e-05,
|
57 |
+
"loss": 0.1614,
|
58 |
+
"step": 70
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.47904191616766467,
|
62 |
+
"grad_norm": 0.8221730589866638,
|
63 |
+
"learning_rate": 9.580838323353294e-05,
|
64 |
+
"loss": 0.1408,
|
65 |
+
"step": 80
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.5389221556886228,
|
69 |
+
"grad_norm": 0.6895884871482849,
|
70 |
+
"learning_rate": 0.00010778443113772456,
|
71 |
+
"loss": 0.1253,
|
72 |
+
"step": 90
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.5988023952095808,
|
76 |
+
"grad_norm": 1.0663549900054932,
|
77 |
+
"learning_rate": 0.00011976047904191617,
|
78 |
+
"loss": 0.1049,
|
79 |
+
"step": 100
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.6586826347305389,
|
83 |
+
"grad_norm": 0.4257476031780243,
|
84 |
+
"learning_rate": 0.0001317365269461078,
|
85 |
+
"loss": 0.1017,
|
86 |
+
"step": 110
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.718562874251497,
|
90 |
+
"grad_norm": 0.46928706765174866,
|
91 |
+
"learning_rate": 0.0001437125748502994,
|
92 |
+
"loss": 0.0887,
|
93 |
+
"step": 120
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.7784431137724551,
|
97 |
+
"grad_norm": 0.40188413858413696,
|
98 |
+
"learning_rate": 0.00015568862275449103,
|
99 |
+
"loss": 0.0821,
|
100 |
+
"step": 130
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.8383233532934131,
|
104 |
+
"grad_norm": 0.44656357169151306,
|
105 |
+
"learning_rate": 0.00016766467065868263,
|
106 |
+
"loss": 0.0802,
|
107 |
+
"step": 140
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.8982035928143712,
|
111 |
+
"grad_norm": 0.8075553178787231,
|
112 |
+
"learning_rate": 0.00017964071856287425,
|
113 |
+
"loss": 0.0768,
|
114 |
+
"step": 150
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.9580838323353293,
|
118 |
+
"grad_norm": 0.7961329817771912,
|
119 |
+
"learning_rate": 0.00019161676646706587,
|
120 |
+
"loss": 0.0788,
|
121 |
+
"step": 160
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 1.0179640718562875,
|
125 |
+
"grad_norm": 0.5878146886825562,
|
126 |
+
"learning_rate": 0.00019999955550902674,
|
127 |
+
"loss": 0.0751,
|
128 |
+
"step": 170
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 1.0778443113772456,
|
132 |
+
"grad_norm": 0.4572954773902893,
|
133 |
+
"learning_rate": 0.0001999916535572039,
|
134 |
+
"loss": 0.0684,
|
135 |
+
"step": 180
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 1.1377245508982037,
|
139 |
+
"grad_norm": 0.3957609236240387,
|
140 |
+
"learning_rate": 0.00019997387492660495,
|
141 |
+
"loss": 0.0651,
|
142 |
+
"step": 190
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 1.1976047904191618,
|
146 |
+
"grad_norm": 0.3396591544151306,
|
147 |
+
"learning_rate": 0.00019994622137331462,
|
148 |
+
"loss": 0.0655,
|
149 |
+
"step": 200
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 1.2574850299401197,
|
153 |
+
"grad_norm": 0.4531398415565491,
|
154 |
+
"learning_rate": 0.00019990869562881356,
|
155 |
+
"loss": 0.0648,
|
156 |
+
"step": 210
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 1.3173652694610778,
|
160 |
+
"grad_norm": 0.36353209614753723,
|
161 |
+
"learning_rate": 0.00019986130139970835,
|
162 |
+
"loss": 0.0619,
|
163 |
+
"step": 220
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 1.377245508982036,
|
167 |
+
"grad_norm": 0.392230361700058,
|
168 |
+
"learning_rate": 0.00019980404336736554,
|
169 |
+
"loss": 0.0583,
|
170 |
+
"step": 230
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 1.437125748502994,
|
174 |
+
"grad_norm": 0.34288108348846436,
|
175 |
+
"learning_rate": 0.00019973692718744902,
|
176 |
+
"loss": 0.0599,
|
177 |
+
"step": 240
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 1.4970059880239521,
|
181 |
+
"grad_norm": 0.40226370096206665,
|
182 |
+
"learning_rate": 0.0001996599594893617,
|
183 |
+
"loss": 0.0538,
|
184 |
+
"step": 250
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 1.55688622754491,
|
188 |
+
"grad_norm": 0.26641878485679626,
|
189 |
+
"learning_rate": 0.00019957314787559036,
|
190 |
+
"loss": 0.0557,
|
191 |
+
"step": 260
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 1.6167664670658684,
|
195 |
+
"grad_norm": 0.4017570912837982,
|
196 |
+
"learning_rate": 0.00019947650092095494,
|
197 |
+
"loss": 0.0514,
|
198 |
+
"step": 270
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 1.6766467065868262,
|
202 |
+
"grad_norm": 0.2945254445075989,
|
203 |
+
"learning_rate": 0.00019937002817176146,
|
204 |
+
"loss": 0.0506,
|
205 |
+
"step": 280
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 1.7365269461077846,
|
209 |
+
"grad_norm": 0.2502083480358124,
|
210 |
+
"learning_rate": 0.00019925374014485912,
|
211 |
+
"loss": 0.0518,
|
212 |
+
"step": 290
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 1.7964071856287425,
|
216 |
+
"grad_norm": 0.3017183840274811,
|
217 |
+
"learning_rate": 0.00019912764832660156,
|
218 |
+
"loss": 0.0541,
|
219 |
+
"step": 300
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 1.8562874251497006,
|
223 |
+
"grad_norm": 0.23154252767562866,
|
224 |
+
"learning_rate": 0.00019899176517171208,
|
225 |
+
"loss": 0.0501,
|
226 |
+
"step": 310
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 1.9161676646706587,
|
230 |
+
"grad_norm": 0.2425275295972824,
|
231 |
+
"learning_rate": 0.0001988461041020537,
|
232 |
+
"loss": 0.0437,
|
233 |
+
"step": 320
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 1.9760479041916168,
|
237 |
+
"grad_norm": 0.2978982925415039,
|
238 |
+
"learning_rate": 0.0001986906795053032,
|
239 |
+
"loss": 0.045,
|
240 |
+
"step": 330
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 2.035928143712575,
|
244 |
+
"grad_norm": 0.3514959514141083,
|
245 |
+
"learning_rate": 0.00019852550673353004,
|
246 |
+
"loss": 0.0492,
|
247 |
+
"step": 340
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 2.095808383233533,
|
251 |
+
"grad_norm": 0.488469660282135,
|
252 |
+
"learning_rate": 0.00019835060210168017,
|
253 |
+
"loss": 0.0479,
|
254 |
+
"step": 350
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 2.155688622754491,
|
258 |
+
"grad_norm": 0.26303964853286743,
|
259 |
+
"learning_rate": 0.00019816598288596412,
|
260 |
+
"loss": 0.0475,
|
261 |
+
"step": 360
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 2.215568862275449,
|
265 |
+
"grad_norm": 0.25606536865234375,
|
266 |
+
"learning_rate": 0.00019797166732215076,
|
267 |
+
"loss": 0.0413,
|
268 |
+
"step": 370
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 2.2754491017964074,
|
272 |
+
"grad_norm": 0.291185200214386,
|
273 |
+
"learning_rate": 0.0001977676746037662,
|
274 |
+
"loss": 0.0443,
|
275 |
+
"step": 380
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 2.3353293413173652,
|
279 |
+
"grad_norm": 0.314779669046402,
|
280 |
+
"learning_rate": 0.0001975540248801977,
|
281 |
+
"loss": 0.044,
|
282 |
+
"step": 390
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 2.3952095808383236,
|
286 |
+
"grad_norm": 0.259378582239151,
|
287 |
+
"learning_rate": 0.00019733073925470352,
|
288 |
+
"loss": 0.0413,
|
289 |
+
"step": 400
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 2.4550898203592815,
|
293 |
+
"grad_norm": 0.5142641067504883,
|
294 |
+
"learning_rate": 0.00019709783978232842,
|
295 |
+
"loss": 0.0453,
|
296 |
+
"step": 410
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 2.5149700598802394,
|
300 |
+
"grad_norm": 0.33504176139831543,
|
301 |
+
"learning_rate": 0.00019685534946772536,
|
302 |
+
"loss": 0.0438,
|
303 |
+
"step": 420
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 2.5748502994011977,
|
307 |
+
"grad_norm": 0.33974334597587585,
|
308 |
+
"learning_rate": 0.00019660329226288273,
|
309 |
+
"loss": 0.0436,
|
310 |
+
"step": 430
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 2.6347305389221556,
|
314 |
+
"grad_norm": 0.29948851466178894,
|
315 |
+
"learning_rate": 0.00019634169306475917,
|
316 |
+
"loss": 0.0404,
|
317 |
+
"step": 440
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 2.694610778443114,
|
321 |
+
"grad_norm": 0.2677200436592102,
|
322 |
+
"learning_rate": 0.00019607057771282372,
|
323 |
+
"loss": 0.0423,
|
324 |
+
"step": 450
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 2.754491017964072,
|
328 |
+
"grad_norm": 0.2578097879886627,
|
329 |
+
"learning_rate": 0.00019578997298650395,
|
330 |
+
"loss": 0.0385,
|
331 |
+
"step": 460
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 2.81437125748503,
|
335 |
+
"grad_norm": 0.2699326276779175,
|
336 |
+
"learning_rate": 0.00019549990660254073,
|
337 |
+
"loss": 0.0402,
|
338 |
+
"step": 470
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 2.874251497005988,
|
342 |
+
"grad_norm": 0.22369785606861115,
|
343 |
+
"learning_rate": 0.00019520040721225033,
|
344 |
+
"loss": 0.0402,
|
345 |
+
"step": 480
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 2.934131736526946,
|
349 |
+
"grad_norm": 0.23244518041610718,
|
350 |
+
"learning_rate": 0.00019489150439869465,
|
351 |
+
"loss": 0.0393,
|
352 |
+
"step": 490
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 2.9940119760479043,
|
356 |
+
"grad_norm": 0.2809353768825531,
|
357 |
+
"learning_rate": 0.000194573228673759,
|
358 |
+
"loss": 0.0402,
|
359 |
+
"step": 500
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 3.053892215568862,
|
363 |
+
"grad_norm": 0.1485118865966797,
|
364 |
+
"learning_rate": 0.0001942456114751382,
|
365 |
+
"loss": 0.0364,
|
366 |
+
"step": 510
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 3.1137724550898205,
|
370 |
+
"grad_norm": 0.35376912355422974,
|
371 |
+
"learning_rate": 0.00019390868516323156,
|
372 |
+
"loss": 0.0358,
|
373 |
+
"step": 520
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 3.1736526946107784,
|
377 |
+
"grad_norm": 0.24453511834144592,
|
378 |
+
"learning_rate": 0.00019356248301794632,
|
379 |
+
"loss": 0.0366,
|
380 |
+
"step": 530
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 3.2335329341317367,
|
384 |
+
"grad_norm": 0.2331007570028305,
|
385 |
+
"learning_rate": 0.00019320703923541045,
|
386 |
+
"loss": 0.0387,
|
387 |
+
"step": 540
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 3.2934131736526946,
|
391 |
+
"grad_norm": 0.2162063866853714,
|
392 |
+
"learning_rate": 0.00019284238892459485,
|
393 |
+
"loss": 0.0373,
|
394 |
+
"step": 550
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 3.3532934131736525,
|
398 |
+
"grad_norm": 0.22156478464603424,
|
399 |
+
"learning_rate": 0.00019246856810384576,
|
400 |
+
"loss": 0.042,
|
401 |
+
"step": 560
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 3.413173652694611,
|
405 |
+
"grad_norm": 0.31770089268684387,
|
406 |
+
"learning_rate": 0.0001920856136973266,
|
407 |
+
"loss": 0.0421,
|
408 |
+
"step": 570
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 3.4730538922155687,
|
412 |
+
"grad_norm": 0.37347620725631714,
|
413 |
+
"learning_rate": 0.0001916935635313711,
|
414 |
+
"loss": 0.0354,
|
415 |
+
"step": 580
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 3.532934131736527,
|
419 |
+
"grad_norm": 0.2527390420436859,
|
420 |
+
"learning_rate": 0.00019129245633074703,
|
421 |
+
"loss": 0.0378,
|
422 |
+
"step": 590
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 3.592814371257485,
|
426 |
+
"grad_norm": 0.2797776162624359,
|
427 |
+
"learning_rate": 0.0001908823317148309,
|
428 |
+
"loss": 0.039,
|
429 |
+
"step": 600
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 3.6526946107784433,
|
433 |
+
"grad_norm": 0.3039669692516327,
|
434 |
+
"learning_rate": 0.00019046323019369474,
|
435 |
+
"loss": 0.0376,
|
436 |
+
"step": 610
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 3.712574850299401,
|
440 |
+
"grad_norm": 0.3194757103919983,
|
441 |
+
"learning_rate": 0.0001900351931641047,
|
442 |
+
"loss": 0.0364,
|
443 |
+
"step": 620
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 3.772455089820359,
|
447 |
+
"grad_norm": 0.29081788659095764,
|
448 |
+
"learning_rate": 0.00018959826290543207,
|
449 |
+
"loss": 0.0387,
|
450 |
+
"step": 630
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 3.8323353293413174,
|
454 |
+
"grad_norm": 0.2967931628227234,
|
455 |
+
"learning_rate": 0.00018915248257547713,
|
456 |
+
"loss": 0.0363,
|
457 |
+
"step": 640
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 3.8922155688622757,
|
461 |
+
"grad_norm": 0.29823440313339233,
|
462 |
+
"learning_rate": 0.0001886978962062062,
|
463 |
+
"loss": 0.0413,
|
464 |
+
"step": 650
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 3.9520958083832336,
|
468 |
+
"grad_norm": 0.2680460810661316,
|
469 |
+
"learning_rate": 0.00018823454869940242,
|
470 |
+
"loss": 0.0381,
|
471 |
+
"step": 660
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 4.0119760479041915,
|
475 |
+
"grad_norm": 0.18794329464435577,
|
476 |
+
"learning_rate": 0.00018776248582223063,
|
477 |
+
"loss": 0.035,
|
478 |
+
"step": 670
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 4.07185628742515,
|
482 |
+
"grad_norm": 0.31890541315078735,
|
483 |
+
"learning_rate": 0.00018728175420271658,
|
484 |
+
"loss": 0.0309,
|
485 |
+
"step": 680
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 4.131736526946108,
|
489 |
+
"grad_norm": 0.2308373749256134,
|
490 |
+
"learning_rate": 0.00018679240132514136,
|
491 |
+
"loss": 0.0338,
|
492 |
+
"step": 690
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 4.191616766467066,
|
496 |
+
"grad_norm": 0.21351340413093567,
|
497 |
+
"learning_rate": 0.0001862944755253511,
|
498 |
+
"loss": 0.0319,
|
499 |
+
"step": 700
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 4.251497005988024,
|
503 |
+
"grad_norm": 0.22991211712360382,
|
504 |
+
"learning_rate": 0.00018578802598598256,
|
505 |
+
"loss": 0.0339,
|
506 |
+
"step": 710
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 4.311377245508982,
|
510 |
+
"grad_norm": 0.1862100064754486,
|
511 |
+
"learning_rate": 0.00018527310273160522,
|
512 |
+
"loss": 0.0323,
|
513 |
+
"step": 720
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 4.37125748502994,
|
517 |
+
"grad_norm": 0.34272634983062744,
|
518 |
+
"learning_rate": 0.00018474975662377996,
|
519 |
+
"loss": 0.0324,
|
520 |
+
"step": 730
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 4.431137724550898,
|
524 |
+
"grad_norm": 0.3091452419757843,
|
525 |
+
"learning_rate": 0.00018421803935603532,
|
526 |
+
"loss": 0.0338,
|
527 |
+
"step": 740
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 4.491017964071856,
|
531 |
+
"grad_norm": 0.31789442896842957,
|
532 |
+
"learning_rate": 0.0001836780034487614,
|
533 |
+
"loss": 0.0343,
|
534 |
+
"step": 750
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 4.550898203592815,
|
538 |
+
"grad_norm": 0.46529826521873474,
|
539 |
+
"learning_rate": 0.00018312970224402224,
|
540 |
+
"loss": 0.0359,
|
541 |
+
"step": 760
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 4.610778443113772,
|
545 |
+
"grad_norm": 0.3199298679828644,
|
546 |
+
"learning_rate": 0.00018257318990028682,
|
547 |
+
"loss": 0.0375,
|
548 |
+
"step": 770
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 4.6706586826347305,
|
552 |
+
"grad_norm": 0.30249571800231934,
|
553 |
+
"learning_rate": 0.00018200852138707975,
|
554 |
+
"loss": 0.039,
|
555 |
+
"step": 780
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 4.730538922155689,
|
559 |
+
"grad_norm": 0.18588168919086456,
|
560 |
+
"learning_rate": 0.00018143575247955135,
|
561 |
+
"loss": 0.0338,
|
562 |
+
"step": 790
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 4.790419161676647,
|
566 |
+
"grad_norm": 0.2602488100528717,
|
567 |
+
"learning_rate": 0.0001808549397529689,
|
568 |
+
"loss": 0.034,
|
569 |
+
"step": 800
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 4.850299401197605,
|
573 |
+
"grad_norm": 0.3416588306427002,
|
574 |
+
"learning_rate": 0.00018026614057712795,
|
575 |
+
"loss": 0.0321,
|
576 |
+
"step": 810
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 4.910179640718563,
|
580 |
+
"grad_norm": 0.31246596574783325,
|
581 |
+
"learning_rate": 0.00017966941311068594,
|
582 |
+
"loss": 0.0313,
|
583 |
+
"step": 820
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 4.970059880239521,
|
587 |
+
"grad_norm": 0.20728591084480286,
|
588 |
+
"learning_rate": 0.00017906481629541743,
|
589 |
+
"loss": 0.0315,
|
590 |
+
"step": 830
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 5.029940119760479,
|
594 |
+
"grad_norm": 0.2376144826412201,
|
595 |
+
"learning_rate": 0.00017845240985039218,
|
596 |
+
"loss": 0.0301,
|
597 |
+
"step": 840
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 5.089820359281437,
|
601 |
+
"grad_norm": 0.2364414483308792,
|
602 |
+
"learning_rate": 0.0001778322542660764,
|
603 |
+
"loss": 0.0299,
|
604 |
+
"step": 850
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 5.149700598802395,
|
608 |
+
"grad_norm": 0.22445429861545563,
|
609 |
+
"learning_rate": 0.0001772044107983577,
|
610 |
+
"loss": 0.0296,
|
611 |
+
"step": 860
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 5.209580838323353,
|
615 |
+
"grad_norm": 0.2109883427619934,
|
616 |
+
"learning_rate": 0.0001765689414624947,
|
617 |
+
"loss": 0.0326,
|
618 |
+
"step": 870
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 5.269461077844311,
|
622 |
+
"grad_norm": 0.17161297798156738,
|
623 |
+
"learning_rate": 0.00017592590902699138,
|
624 |
+
"loss": 0.0321,
|
625 |
+
"step": 880
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 5.3293413173652695,
|
629 |
+
"grad_norm": 0.18407276272773743,
|
630 |
+
"learning_rate": 0.00017527537700739708,
|
631 |
+
"loss": 0.0339,
|
632 |
+
"step": 890
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 5.389221556886228,
|
636 |
+
"grad_norm": 0.1970146745443344,
|
637 |
+
"learning_rate": 0.00017461740966003282,
|
638 |
+
"loss": 0.0356,
|
639 |
+
"step": 900
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 5.449101796407185,
|
643 |
+
"grad_norm": 0.1647009700536728,
|
644 |
+
"learning_rate": 0.00017395207197564442,
|
645 |
+
"loss": 0.0303,
|
646 |
+
"step": 910
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 5.508982035928144,
|
650 |
+
"grad_norm": 0.17965902388095856,
|
651 |
+
"learning_rate": 0.00017327942967298285,
|
652 |
+
"loss": 0.0273,
|
653 |
+
"step": 920
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 5.568862275449102,
|
657 |
+
"grad_norm": 0.2345510870218277,
|
658 |
+
"learning_rate": 0.0001725995491923131,
|
659 |
+
"loss": 0.0274,
|
660 |
+
"step": 930
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 5.62874251497006,
|
664 |
+
"grad_norm": 0.33345136046409607,
|
665 |
+
"learning_rate": 0.00017191249768885137,
|
666 |
+
"loss": 0.0341,
|
667 |
+
"step": 940
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 5.688622754491018,
|
671 |
+
"grad_norm": 0.1947910636663437,
|
672 |
+
"learning_rate": 0.00017121834302613186,
|
673 |
+
"loss": 0.029,
|
674 |
+
"step": 950
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 5.748502994011976,
|
678 |
+
"grad_norm": 0.17892029881477356,
|
679 |
+
"learning_rate": 0.00017051715376930363,
|
680 |
+
"loss": 0.0343,
|
681 |
+
"step": 960
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 5.808383233532934,
|
685 |
+
"grad_norm": 0.174368754029274,
|
686 |
+
"learning_rate": 0.00016980899917835783,
|
687 |
+
"loss": 0.0299,
|
688 |
+
"step": 970
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 5.868263473053892,
|
692 |
+
"grad_norm": 0.24912439286708832,
|
693 |
+
"learning_rate": 0.00016909394920128695,
|
694 |
+
"loss": 0.0295,
|
695 |
+
"step": 980
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 5.92814371257485,
|
699 |
+
"grad_norm": 0.2845669984817505,
|
700 |
+
"learning_rate": 0.00016837207446717518,
|
701 |
+
"loss": 0.0289,
|
702 |
+
"step": 990
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 5.9880239520958085,
|
706 |
+
"grad_norm": 0.22086086869239807,
|
707 |
+
"learning_rate": 0.00016764344627922254,
|
708 |
+
"loss": 0.0327,
|
709 |
+
"step": 1000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 6.047904191616767,
|
713 |
+
"grad_norm": 0.3092862665653229,
|
714 |
+
"learning_rate": 0.0001669081366077016,
|
715 |
+
"loss": 0.026,
|
716 |
+
"step": 1010
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 6.107784431137724,
|
720 |
+
"grad_norm": 0.3457212746143341,
|
721 |
+
"learning_rate": 0.00016616621808284864,
|
722 |
+
"loss": 0.0291,
|
723 |
+
"step": 1020
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 6.167664670658683,
|
727 |
+
"grad_norm": 0.19416464865207672,
|
728 |
+
"learning_rate": 0.00016541776398768966,
|
729 |
+
"loss": 0.0297,
|
730 |
+
"step": 1030
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 6.227544910179641,
|
734 |
+
"grad_norm": 0.24748080968856812,
|
735 |
+
"learning_rate": 0.00016466284825080184,
|
736 |
+
"loss": 0.028,
|
737 |
+
"step": 1040
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 6.287425149700598,
|
741 |
+
"grad_norm": 0.20852451026439667,
|
742 |
+
"learning_rate": 0.00016390154543901122,
|
743 |
+
"loss": 0.0237,
|
744 |
+
"step": 1050
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 6.347305389221557,
|
748 |
+
"grad_norm": 0.2635754346847534,
|
749 |
+
"learning_rate": 0.00016313393075002736,
|
750 |
+
"loss": 0.0267,
|
751 |
+
"step": 1060
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 6.407185628742515,
|
755 |
+
"grad_norm": 0.28253108263015747,
|
756 |
+
"learning_rate": 0.00016236008000501573,
|
757 |
+
"loss": 0.0257,
|
758 |
+
"step": 1070
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 6.467065868263473,
|
762 |
+
"grad_norm": 0.22715069353580475,
|
763 |
+
"learning_rate": 0.00016158006964110837,
|
764 |
+
"loss": 0.0308,
|
765 |
+
"step": 1080
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 6.526946107784431,
|
769 |
+
"grad_norm": 0.2551535964012146,
|
770 |
+
"learning_rate": 0.0001607939767038539,
|
771 |
+
"loss": 0.0265,
|
772 |
+
"step": 1090
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 6.586826347305389,
|
776 |
+
"grad_norm": 0.1504376381635666,
|
777 |
+
"learning_rate": 0.00016000187883960734,
|
778 |
+
"loss": 0.0264,
|
779 |
+
"step": 1100
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 6.6467065868263475,
|
783 |
+
"grad_norm": 0.3545721173286438,
|
784 |
+
"learning_rate": 0.00015920385428786057,
|
785 |
+
"loss": 0.026,
|
786 |
+
"step": 1110
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 6.706586826347305,
|
790 |
+
"grad_norm": 0.16215580701828003,
|
791 |
+
"learning_rate": 0.0001583999818735142,
|
792 |
+
"loss": 0.024,
|
793 |
+
"step": 1120
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 6.766467065868263,
|
797 |
+
"grad_norm": 0.24395804107189178,
|
798 |
+
"learning_rate": 0.00015759034099909158,
|
799 |
+
"loss": 0.0275,
|
800 |
+
"step": 1130
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 6.826347305389222,
|
804 |
+
"grad_norm": 0.18488508462905884,
|
805 |
+
"learning_rate": 0.00015677501163689604,
|
806 |
+
"loss": 0.0299,
|
807 |
+
"step": 1140
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 6.88622754491018,
|
811 |
+
"grad_norm": 0.1932033747434616,
|
812 |
+
"learning_rate": 0.00015595407432111134,
|
813 |
+
"loss": 0.0259,
|
814 |
+
"step": 1150
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 6.946107784431137,
|
818 |
+
"grad_norm": 0.19903336465358734,
|
819 |
+
"learning_rate": 0.00015512761013984711,
|
820 |
+
"loss": 0.0278,
|
821 |
+
"step": 1160
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 7.005988023952096,
|
825 |
+
"grad_norm": 0.24045361578464508,
|
826 |
+
"learning_rate": 0.00015429570072712925,
|
827 |
+
"loss": 0.029,
|
828 |
+
"step": 1170
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 7.065868263473054,
|
832 |
+
"grad_norm": 0.24332760274410248,
|
833 |
+
"learning_rate": 0.00015345842825483666,
|
834 |
+
"loss": 0.0271,
|
835 |
+
"step": 1180
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 7.125748502994012,
|
839 |
+
"grad_norm": 0.2306133508682251,
|
840 |
+
"learning_rate": 0.00015261587542458443,
|
841 |
+
"loss": 0.0232,
|
842 |
+
"step": 1190
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 7.18562874251497,
|
846 |
+
"grad_norm": 0.2224716693162918,
|
847 |
+
"learning_rate": 0.0001517681254595554,
|
848 |
+
"loss": 0.0246,
|
849 |
+
"step": 1200
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 7.245508982035928,
|
853 |
+
"grad_norm": 0.2902887463569641,
|
854 |
+
"learning_rate": 0.00015091526209627944,
|
855 |
+
"loss": 0.024,
|
856 |
+
"step": 1210
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 7.3053892215568865,
|
860 |
+
"grad_norm": 0.35143372416496277,
|
861 |
+
"learning_rate": 0.0001500573695763625,
|
862 |
+
"loss": 0.0248,
|
863 |
+
"step": 1220
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 7.365269461077844,
|
867 |
+
"grad_norm": 0.3402245342731476,
|
868 |
+
"learning_rate": 0.00014919453263816565,
|
869 |
+
"loss": 0.0256,
|
870 |
+
"step": 1230
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 7.425149700598802,
|
874 |
+
"grad_norm": 0.17631883919239044,
|
875 |
+
"learning_rate": 0.0001483268365084351,
|
876 |
+
"loss": 0.0274,
|
877 |
+
"step": 1240
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 7.485029940119761,
|
881 |
+
"grad_norm": 0.1852467656135559,
|
882 |
+
"learning_rate": 0.0001474543668938837,
|
883 |
+
"loss": 0.0242,
|
884 |
+
"step": 1250
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 7.544910179640718,
|
888 |
+
"grad_norm": 0.20840848982334137,
|
889 |
+
"learning_rate": 0.0001465772099727256,
|
890 |
+
"loss": 0.027,
|
891 |
+
"step": 1260
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 7.604790419161676,
|
895 |
+
"grad_norm": 0.19811859726905823,
|
896 |
+
"learning_rate": 0.00014569545238616368,
|
897 |
+
"loss": 0.0229,
|
898 |
+
"step": 1270
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 7.664670658682635,
|
902 |
+
"grad_norm": 0.15369650721549988,
|
903 |
+
"learning_rate": 0.00014480918122983174,
|
904 |
+
"loss": 0.0266,
|
905 |
+
"step": 1280
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 7.724550898203593,
|
909 |
+
"grad_norm": 0.4044421315193176,
|
910 |
+
"learning_rate": 0.00014391848404519158,
|
911 |
+
"loss": 0.0251,
|
912 |
+
"step": 1290
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 7.7844311377245505,
|
916 |
+
"grad_norm": 0.17981642484664917,
|
917 |
+
"learning_rate": 0.0001430234488108861,
|
918 |
+
"loss": 0.0248,
|
919 |
+
"step": 1300
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 7.844311377245509,
|
923 |
+
"grad_norm": 0.2915557324886322,
|
924 |
+
"learning_rate": 0.00014212416393404915,
|
925 |
+
"loss": 0.0231,
|
926 |
+
"step": 1310
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 7.904191616766467,
|
930 |
+
"grad_norm": 0.20533527433872223,
|
931 |
+
"learning_rate": 0.00014122071824157318,
|
932 |
+
"loss": 0.0257,
|
933 |
+
"step": 1320
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 7.9640718562874255,
|
937 |
+
"grad_norm": 0.15892498195171356,
|
938 |
+
"learning_rate": 0.0001403132009713354,
|
939 |
+
"loss": 0.0253,
|
940 |
+
"step": 1330
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 8.023952095808383,
|
944 |
+
"grad_norm": 0.34935975074768066,
|
945 |
+
"learning_rate": 0.0001394017017633831,
|
946 |
+
"loss": 0.0261,
|
947 |
+
"step": 1340
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 8.08383233532934,
|
951 |
+
"grad_norm": 0.1838160902261734,
|
952 |
+
"learning_rate": 0.00013848631065107973,
|
953 |
+
"loss": 0.0231,
|
954 |
+
"step": 1350
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 8.1437125748503,
|
958 |
+
"grad_norm": 0.24909768998622894,
|
959 |
+
"learning_rate": 0.00013756711805221168,
|
960 |
+
"loss": 0.0256,
|
961 |
+
"step": 1360
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 8.203592814371257,
|
965 |
+
"grad_norm": 0.2749064266681671,
|
966 |
+
"learning_rate": 0.00013664421476005723,
|
967 |
+
"loss": 0.0258,
|
968 |
+
"step": 1370
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 8.263473053892216,
|
972 |
+
"grad_norm": 0.24632593989372253,
|
973 |
+
"learning_rate": 0.0001357176919344186,
|
974 |
+
"loss": 0.023,
|
975 |
+
"step": 1380
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 8.323353293413174,
|
979 |
+
"grad_norm": 0.22254560887813568,
|
980 |
+
"learning_rate": 0.0001347876410926175,
|
981 |
+
"loss": 0.0249,
|
982 |
+
"step": 1390
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 8.383233532934131,
|
986 |
+
"grad_norm": 0.1739770621061325,
|
987 |
+
"learning_rate": 0.00013385415410045555,
|
988 |
+
"loss": 0.0215,
|
989 |
+
"step": 1400
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 8.44311377245509,
|
993 |
+
"grad_norm": 0.15595784783363342,
|
994 |
+
"learning_rate": 0.00013291732316314024,
|
995 |
+
"loss": 0.023,
|
996 |
+
"step": 1410
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 8.502994011976048,
|
1000 |
+
"grad_norm": 0.12900565564632416,
|
1001 |
+
"learning_rate": 0.00013197724081617736,
|
1002 |
+
"loss": 0.0227,
|
1003 |
+
"step": 1420
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 8.562874251497005,
|
1007 |
+
"grad_norm": 0.18408159911632538,
|
1008 |
+
"learning_rate": 0.00013103399991623078,
|
1009 |
+
"loss": 0.0249,
|
1010 |
+
"step": 1430
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 8.622754491017965,
|
1014 |
+
"grad_norm": 0.3930126130580902,
|
1015 |
+
"learning_rate": 0.00013008769363195052,
|
1016 |
+
"loss": 0.0235,
|
1017 |
+
"step": 1440
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 8.682634730538922,
|
1021 |
+
"grad_norm": 0.22931550443172455,
|
1022 |
+
"learning_rate": 0.00012913841543477005,
|
1023 |
+
"loss": 0.02,
|
1024 |
+
"step": 1450
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 8.74251497005988,
|
1028 |
+
"grad_norm": 0.1572323441505432,
|
1029 |
+
"learning_rate": 0.00012818625908967367,
|
1030 |
+
"loss": 0.0244,
|
1031 |
+
"step": 1460
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 8.802395209580839,
|
1035 |
+
"grad_norm": 0.09360220283269882,
|
1036 |
+
"learning_rate": 0.0001272313186459348,
|
1037 |
+
"loss": 0.021,
|
1038 |
+
"step": 1470
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 8.862275449101796,
|
1042 |
+
"grad_norm": 0.17760924994945526,
|
1043 |
+
"learning_rate": 0.00012627368842782626,
|
1044 |
+
"loss": 0.0221,
|
1045 |
+
"step": 1480
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 8.922155688622755,
|
1049 |
+
"grad_norm": 0.18415001034736633,
|
1050 |
+
"learning_rate": 0.00012531346302530348,
|
1051 |
+
"loss": 0.0255,
|
1052 |
+
"step": 1490
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 8.982035928143713,
|
1056 |
+
"grad_norm": 0.479499489068985,
|
1057 |
+
"learning_rate": 0.0001243507372846613,
|
1058 |
+
"loss": 0.0207,
|
1059 |
+
"step": 1500
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 9.04191616766467,
|
1063 |
+
"grad_norm": 0.22728319466114044,
|
1064 |
+
"learning_rate": 0.00012338560629916544,
|
1065 |
+
"loss": 0.0253,
|
1066 |
+
"step": 1510
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 9.10179640718563,
|
1070 |
+
"grad_norm": 0.1806812733411789,
|
1071 |
+
"learning_rate": 0.00012241816539965996,
|
1072 |
+
"loss": 0.0255,
|
1073 |
+
"step": 1520
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 9.161676646706587,
|
1077 |
+
"grad_norm": 0.26954516768455505,
|
1078 |
+
"learning_rate": 0.00012144851014515055,
|
1079 |
+
"loss": 0.0219,
|
1080 |
+
"step": 1530
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 9.221556886227544,
|
1084 |
+
"grad_norm": 0.30244171619415283,
|
1085 |
+
"learning_rate": 0.000120476736313366,
|
1086 |
+
"loss": 0.0253,
|
1087 |
+
"step": 1540
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 9.281437125748504,
|
1091 |
+
"grad_norm": 0.1826360821723938,
|
1092 |
+
"learning_rate": 0.0001195029398912976,
|
1093 |
+
"loss": 0.022,
|
1094 |
+
"step": 1550
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 9.341317365269461,
|
1098 |
+
"grad_norm": 0.22383596003055573,
|
1099 |
+
"learning_rate": 0.00011852721706571803,
|
1100 |
+
"loss": 0.0253,
|
1101 |
+
"step": 1560
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 9.401197604790418,
|
1105 |
+
"grad_norm": 0.1451132595539093,
|
1106 |
+
"learning_rate": 0.00011754966421368066,
|
1107 |
+
"loss": 0.0227,
|
1108 |
+
"step": 1570
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 9.461077844311378,
|
1112 |
+
"grad_norm": 0.1684962958097458,
|
1113 |
+
"learning_rate": 0.00011657037789299961,
|
1114 |
+
"loss": 0.022,
|
1115 |
+
"step": 1580
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 9.520958083832335,
|
1119 |
+
"grad_norm": 0.18602758646011353,
|
1120 |
+
"learning_rate": 0.00011558945483271253,
|
1121 |
+
"loss": 0.0238,
|
1122 |
+
"step": 1590
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 9.580838323353294,
|
1126 |
+
"grad_norm": 0.1612151712179184,
|
1127 |
+
"learning_rate": 0.00011460699192352608,
|
1128 |
+
"loss": 0.0229,
|
1129 |
+
"step": 1600
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 9.640718562874252,
|
1133 |
+
"grad_norm": 0.1520804762840271,
|
1134 |
+
"learning_rate": 0.00011362308620824549,
|
1135 |
+
"loss": 0.0215,
|
1136 |
+
"step": 1610
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 9.70059880239521,
|
1140 |
+
"grad_norm": 0.23648953437805176,
|
1141 |
+
"learning_rate": 0.00011263783487218924,
|
1142 |
+
"loss": 0.0205,
|
1143 |
+
"step": 1620
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 9.760479041916168,
|
1147 |
+
"grad_norm": 0.17972023785114288,
|
1148 |
+
"learning_rate": 0.00011165133523358955,
|
1149 |
+
"loss": 0.0171,
|
1150 |
+
"step": 1630
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 9.820359281437126,
|
1154 |
+
"grad_norm": 0.20409125089645386,
|
1155 |
+
"learning_rate": 0.00011066368473397977,
|
1156 |
+
"loss": 0.0229,
|
1157 |
+
"step": 1640
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 9.880239520958083,
|
1161 |
+
"grad_norm": 0.1429627388715744,
|
1162 |
+
"learning_rate": 0.0001096749809285695,
|
1163 |
+
"loss": 0.0212,
|
1164 |
+
"step": 1650
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 9.940119760479043,
|
1168 |
+
"grad_norm": 0.20382557809352875,
|
1169 |
+
"learning_rate": 0.0001086853214766088,
|
1170 |
+
"loss": 0.019,
|
1171 |
+
"step": 1660
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 10.0,
|
1175 |
+
"grad_norm": 0.25044798851013184,
|
1176 |
+
"learning_rate": 0.00010769480413174162,
|
1177 |
+
"loss": 0.0192,
|
1178 |
+
"step": 1670
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 10.059880239520957,
|
1182 |
+
"grad_norm": 0.18214142322540283,
|
1183 |
+
"learning_rate": 0.0001067035267323504,
|
1184 |
+
"loss": 0.0219,
|
1185 |
+
"step": 1680
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 10.119760479041917,
|
1189 |
+
"grad_norm": 0.17343035340309143,
|
1190 |
+
"learning_rate": 0.00010571158719189192,
|
1191 |
+
"loss": 0.0224,
|
1192 |
+
"step": 1690
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 10.179640718562874,
|
1196 |
+
"grad_norm": 0.1327686607837677,
|
1197 |
+
"learning_rate": 0.00010471908348922596,
|
1198 |
+
"loss": 0.0221,
|
1199 |
+
"step": 1700
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 10.239520958083832,
|
1203 |
+
"grad_norm": 0.22235077619552612,
|
1204 |
+
"learning_rate": 0.00010372611365893758,
|
1205 |
+
"loss": 0.0234,
|
1206 |
+
"step": 1710
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 10.29940119760479,
|
1210 |
+
"grad_norm": 0.2747216522693634,
|
1211 |
+
"learning_rate": 0.00010273277578165341,
|
1212 |
+
"loss": 0.0222,
|
1213 |
+
"step": 1720
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 10.359281437125748,
|
1217 |
+
"grad_norm": 0.22264912724494934,
|
1218 |
+
"learning_rate": 0.00010173916797435405,
|
1219 |
+
"loss": 0.0211,
|
1220 |
+
"step": 1730
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 10.419161676646706,
|
1224 |
+
"grad_norm": 0.6474366784095764,
|
1225 |
+
"learning_rate": 0.00010074538838068243,
|
1226 |
+
"loss": 0.0212,
|
1227 |
+
"step": 1740
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 10.479041916167665,
|
1231 |
+
"grad_norm": 0.24274206161499023,
|
1232 |
+
"learning_rate": 9.975153516124972e-05,
|
1233 |
+
"loss": 0.0214,
|
1234 |
+
"step": 1750
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 10.538922155688622,
|
1238 |
+
"grad_norm": 0.19830718636512756,
|
1239 |
+
"learning_rate": 9.875770648393936e-05,
|
1240 |
+
"loss": 0.0233,
|
1241 |
+
"step": 1760
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 10.598802395209582,
|
1245 |
+
"grad_norm": 0.12125121057033539,
|
1246 |
+
"learning_rate": 9.776400051421078e-05,
|
1247 |
+
"loss": 0.0194,
|
1248 |
+
"step": 1770
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 10.658682634730539,
|
1252 |
+
"grad_norm": 0.16296449303627014,
|
1253 |
+
"learning_rate": 9.677051540540286e-05,
|
1254 |
+
"loss": 0.0225,
|
1255 |
+
"step": 1780
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 10.718562874251496,
|
1259 |
+
"grad_norm": 0.24490657448768616,
|
1260 |
+
"learning_rate": 9.577734928903899e-05,
|
1261 |
+
"loss": 0.0213,
|
1262 |
+
"step": 1790
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 10.778443113772456,
|
1266 |
+
"grad_norm": 0.16158710420131683,
|
1267 |
+
"learning_rate": 9.47846002651341e-05,
|
1268 |
+
"loss": 0.0199,
|
1269 |
+
"step": 1800
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 10.838323353293413,
|
1273 |
+
"grad_norm": 0.13847924768924713,
|
1274 |
+
"learning_rate": 9.379236639250473e-05,
|
1275 |
+
"loss": 0.0185,
|
1276 |
+
"step": 1810
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 10.89820359281437,
|
1280 |
+
"grad_norm": 0.18506354093551636,
|
1281 |
+
"learning_rate": 9.28007456790833e-05,
|
1282 |
+
"loss": 0.0207,
|
1283 |
+
"step": 1820
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 10.95808383233533,
|
1287 |
+
"grad_norm": 0.1724933683872223,
|
1288 |
+
"learning_rate": 9.180983607223757e-05,
|
1289 |
+
"loss": 0.022,
|
1290 |
+
"step": 1830
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 11.017964071856287,
|
1294 |
+
"grad_norm": 0.2975943088531494,
|
1295 |
+
"learning_rate": 9.081973544909557e-05,
|
1296 |
+
"loss": 0.0179,
|
1297 |
+
"step": 1840
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 11.077844311377245,
|
1301 |
+
"grad_norm": 0.12449675798416138,
|
1302 |
+
"learning_rate": 8.983054160687801e-05,
|
1303 |
+
"loss": 0.0211,
|
1304 |
+
"step": 1850
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 11.137724550898204,
|
1308 |
+
"grad_norm": 0.15286274254322052,
|
1309 |
+
"learning_rate": 8.88423522532383e-05,
|
1310 |
+
"loss": 0.0212,
|
1311 |
+
"step": 1860
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 11.197604790419161,
|
1315 |
+
"grad_norm": 0.17604878544807434,
|
1316 |
+
"learning_rate": 8.785526499661137e-05,
|
1317 |
+
"loss": 0.02,
|
1318 |
+
"step": 1870
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 11.25748502994012,
|
1322 |
+
"grad_norm": 0.33563774824142456,
|
1323 |
+
"learning_rate": 8.686937733657264e-05,
|
1324 |
+
"loss": 0.0222,
|
1325 |
+
"step": 1880
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 11.317365269461078,
|
1329 |
+
"grad_norm": 0.1683143824338913,
|
1330 |
+
"learning_rate": 8.588478665420717e-05,
|
1331 |
+
"loss": 0.0215,
|
1332 |
+
"step": 1890
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 11.377245508982035,
|
1336 |
+
"grad_norm": 0.36611148715019226,
|
1337 |
+
"learning_rate": 8.490159020249107e-05,
|
1338 |
+
"loss": 0.0203,
|
1339 |
+
"step": 1900
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 11.437125748502995,
|
1343 |
+
"grad_norm": 0.13477054238319397,
|
1344 |
+
"learning_rate": 8.391988509668534e-05,
|
1345 |
+
"loss": 0.0204,
|
1346 |
+
"step": 1910
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 11.497005988023952,
|
1350 |
+
"grad_norm": 0.25305667519569397,
|
1351 |
+
"learning_rate": 8.293976830474312e-05,
|
1352 |
+
"loss": 0.0222,
|
1353 |
+
"step": 1920
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 11.55688622754491,
|
1357 |
+
"grad_norm": 0.13116030395030975,
|
1358 |
+
"learning_rate": 8.196133663773191e-05,
|
1359 |
+
"loss": 0.0173,
|
1360 |
+
"step": 1930
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 11.616766467065869,
|
1364 |
+
"grad_norm": 0.15626917779445648,
|
1365 |
+
"learning_rate": 8.09846867402709e-05,
|
1366 |
+
"loss": 0.0208,
|
1367 |
+
"step": 1940
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 11.676646706586826,
|
1371 |
+
"grad_norm": 0.13971462845802307,
|
1372 |
+
"learning_rate": 8.000991508098492e-05,
|
1373 |
+
"loss": 0.0177,
|
1374 |
+
"step": 1950
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 11.736526946107784,
|
1378 |
+
"grad_norm": 0.26388251781463623,
|
1379 |
+
"learning_rate": 7.903711794297572e-05,
|
1380 |
+
"loss": 0.019,
|
1381 |
+
"step": 1960
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 11.796407185628743,
|
1385 |
+
"grad_norm": 0.17708484828472137,
|
1386 |
+
"learning_rate": 7.806639141431178e-05,
|
1387 |
+
"loss": 0.0189,
|
1388 |
+
"step": 1970
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 11.8562874251497,
|
1392 |
+
"grad_norm": 0.2096002697944641,
|
1393 |
+
"learning_rate": 7.709783137853692e-05,
|
1394 |
+
"loss": 0.0204,
|
1395 |
+
"step": 1980
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 11.91616766467066,
|
1399 |
+
"grad_norm": 0.14451850950717926,
|
1400 |
+
"learning_rate": 7.613153350519969e-05,
|
1401 |
+
"loss": 0.0173,
|
1402 |
+
"step": 1990
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 11.976047904191617,
|
1406 |
+
"grad_norm": 0.12203036993741989,
|
1407 |
+
"learning_rate": 7.516759324040338e-05,
|
1408 |
+
"loss": 0.0168,
|
1409 |
+
"step": 2000
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 12.035928143712574,
|
1413 |
+
"grad_norm": 0.14412252604961395,
|
1414 |
+
"learning_rate": 7.42061057973785e-05,
|
1415 |
+
"loss": 0.0185,
|
1416 |
+
"step": 2010
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 12.095808383233534,
|
1420 |
+
"grad_norm": 0.27144360542297363,
|
1421 |
+
"learning_rate": 7.324716614707793e-05,
|
1422 |
+
"loss": 0.0195,
|
1423 |
+
"step": 2020
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 12.155688622754491,
|
1427 |
+
"grad_norm": 0.15820185840129852,
|
1428 |
+
"learning_rate": 7.229086900879629e-05,
|
1429 |
+
"loss": 0.0191,
|
1430 |
+
"step": 2030
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 12.215568862275449,
|
1434 |
+
"grad_norm": 0.16200768947601318,
|
1435 |
+
"learning_rate": 7.133730884081395e-05,
|
1436 |
+
"loss": 0.0193,
|
1437 |
+
"step": 2040
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 12.275449101796408,
|
1441 |
+
"grad_norm": 0.18649928271770477,
|
1442 |
+
"learning_rate": 7.038657983106702e-05,
|
1443 |
+
"loss": 0.0159,
|
1444 |
+
"step": 2050
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 12.335329341317365,
|
1448 |
+
"grad_norm": 0.17984548211097717,
|
1449 |
+
"learning_rate": 6.943877588784382e-05,
|
1450 |
+
"loss": 0.0199,
|
1451 |
+
"step": 2060
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 12.395209580838323,
|
1455 |
+
"grad_norm": 0.16168512403964996,
|
1456 |
+
"learning_rate": 6.849399063050912e-05,
|
1457 |
+
"loss": 0.0148,
|
1458 |
+
"step": 2070
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 12.455089820359282,
|
1462 |
+
"grad_norm": 0.14747169613838196,
|
1463 |
+
"learning_rate": 6.755231738025693e-05,
|
1464 |
+
"loss": 0.0189,
|
1465 |
+
"step": 2080
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 12.51497005988024,
|
1469 |
+
"grad_norm": 0.1350451409816742,
|
1470 |
+
"learning_rate": 6.661384915089268e-05,
|
1471 |
+
"loss": 0.0142,
|
1472 |
+
"step": 2090
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 12.574850299401197,
|
1476 |
+
"grad_norm": 0.12176797538995743,
|
1477 |
+
"learning_rate": 6.567867863964561e-05,
|
1478 |
+
"loss": 0.018,
|
1479 |
+
"step": 2100
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 12.634730538922156,
|
1483 |
+
"grad_norm": 0.12629830837249756,
|
1484 |
+
"learning_rate": 6.474689821801295e-05,
|
1485 |
+
"loss": 0.0151,
|
1486 |
+
"step": 2110
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 12.694610778443113,
|
1490 |
+
"grad_norm": 0.2103053331375122,
|
1491 |
+
"learning_rate": 6.38185999226356e-05,
|
1492 |
+
"loss": 0.0148,
|
1493 |
+
"step": 2120
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 12.754491017964071,
|
1497 |
+
"grad_norm": 0.18460889160633087,
|
1498 |
+
"learning_rate": 6.289387544620736e-05,
|
1499 |
+
"loss": 0.0172,
|
1500 |
+
"step": 2130
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 12.81437125748503,
|
1504 |
+
"grad_norm": 0.22776320576667786,
|
1505 |
+
"learning_rate": 6.197281612841795e-05,
|
1506 |
+
"loss": 0.0175,
|
1507 |
+
"step": 2140
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 12.874251497005988,
|
1511 |
+
"grad_norm": 0.1890844702720642,
|
1512 |
+
"learning_rate": 6.10555129469309e-05,
|
1513 |
+
"loss": 0.0154,
|
1514 |
+
"step": 2150
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 12.934131736526947,
|
1518 |
+
"grad_norm": 0.17597252130508423,
|
1519 |
+
"learning_rate": 6.014205650839737e-05,
|
1520 |
+
"loss": 0.0151,
|
1521 |
+
"step": 2160
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 12.994011976047904,
|
1525 |
+
"grad_norm": 0.15856590867042542,
|
1526 |
+
"learning_rate": 5.923253703950617e-05,
|
1527 |
+
"loss": 0.0154,
|
1528 |
+
"step": 2170
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 13.053892215568862,
|
1532 |
+
"grad_norm": 0.22898879647254944,
|
1533 |
+
"learning_rate": 5.832704437807195e-05,
|
1534 |
+
"loss": 0.0137,
|
1535 |
+
"step": 2180
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 13.113772455089821,
|
1539 |
+
"grad_norm": 0.34772729873657227,
|
1540 |
+
"learning_rate": 5.742566796416129e-05,
|
1541 |
+
"loss": 0.0144,
|
1542 |
+
"step": 2190
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 13.173652694610778,
|
1546 |
+
"grad_norm": 0.18779855966567993,
|
1547 |
+
"learning_rate": 5.652849683125823e-05,
|
1548 |
+
"loss": 0.0146,
|
1549 |
+
"step": 2200
|
1550 |
+
},
|
1551 |
+
{
|
1552 |
+
"epoch": 13.233532934131736,
|
1553 |
+
"grad_norm": 0.17576445639133453,
|
1554 |
+
"learning_rate": 5.563561959747019e-05,
|
1555 |
+
"loss": 0.0196,
|
1556 |
+
"step": 2210
|
1557 |
+
},
|
1558 |
+
{
|
1559 |
+
"epoch": 13.293413173652695,
|
1560 |
+
"grad_norm": 0.15298061072826385,
|
1561 |
+
"learning_rate": 5.4747124456774445e-05,
|
1562 |
+
"loss": 0.0178,
|
1563 |
+
"step": 2220
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"epoch": 13.353293413173652,
|
1567 |
+
"grad_norm": 0.14338500797748566,
|
1568 |
+
"learning_rate": 5.38630991703069e-05,
|
1569 |
+
"loss": 0.0179,
|
1570 |
+
"step": 2230
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"epoch": 13.41317365269461,
|
1574 |
+
"grad_norm": 0.18957915902137756,
|
1575 |
+
"learning_rate": 5.2983631057693484e-05,
|
1576 |
+
"loss": 0.0165,
|
1577 |
+
"step": 2240
|
1578 |
+
},
|
1579 |
+
{
|
1580 |
+
"epoch": 13.47305389221557,
|
1581 |
+
"grad_norm": 0.19262787699699402,
|
1582 |
+
"learning_rate": 5.210880698842514e-05,
|
1583 |
+
"loss": 0.0139,
|
1584 |
+
"step": 2250
|
1585 |
+
},
|
1586 |
+
{
|
1587 |
+
"epoch": 13.532934131736527,
|
1588 |
+
"grad_norm": 0.3551245331764221,
|
1589 |
+
"learning_rate": 5.123871337327724e-05,
|
1590 |
+
"loss": 0.0172,
|
1591 |
+
"step": 2260
|
1592 |
+
},
|
1593 |
+
{
|
1594 |
+
"epoch": 13.592814371257486,
|
1595 |
+
"grad_norm": 0.21090568602085114,
|
1596 |
+
"learning_rate": 5.037343615577448e-05,
|
1597 |
+
"loss": 0.0159,
|
1598 |
+
"step": 2270
|
1599 |
+
},
|
1600 |
+
{
|
1601 |
+
"epoch": 13.652694610778443,
|
1602 |
+
"grad_norm": 0.10332461446523666,
|
1603 |
+
"learning_rate": 4.951306080370165e-05,
|
1604 |
+
"loss": 0.0147,
|
1605 |
+
"step": 2280
|
1606 |
+
},
|
1607 |
+
{
|
1608 |
+
"epoch": 13.7125748502994,
|
1609 |
+
"grad_norm": 0.12004910409450531,
|
1610 |
+
"learning_rate": 4.865767230066166e-05,
|
1611 |
+
"loss": 0.0132,
|
1612 |
+
"step": 2290
|
1613 |
+
},
|
1614 |
+
{
|
1615 |
+
"epoch": 13.77245508982036,
|
1616 |
+
"grad_norm": 0.17986303567886353,
|
1617 |
+
"learning_rate": 4.78073551376812e-05,
|
1618 |
+
"loss": 0.0142,
|
1619 |
+
"step": 2300
|
1620 |
+
},
|
1621 |
+
{
|
1622 |
+
"epoch": 13.832335329341317,
|
1623 |
+
"grad_norm": 0.17033332586288452,
|
1624 |
+
"learning_rate": 4.6962193304865335e-05,
|
1625 |
+
"loss": 0.0145,
|
1626 |
+
"step": 2310
|
1627 |
+
},
|
1628 |
+
{
|
1629 |
+
"epoch": 13.892215568862275,
|
1630 |
+
"grad_norm": 0.5156183838844299,
|
1631 |
+
"learning_rate": 4.6122270283100944e-05,
|
1632 |
+
"loss": 0.0162,
|
1633 |
+
"step": 2320
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"epoch": 13.952095808383234,
|
1637 |
+
"grad_norm": 0.47306326031684875,
|
1638 |
+
"learning_rate": 4.528766903581149e-05,
|
1639 |
+
"loss": 0.0151,
|
1640 |
+
"step": 2330
|
1641 |
+
},
|
1642 |
+
{
|
1643 |
+
"epoch": 14.011976047904191,
|
1644 |
+
"grad_norm": 0.10242114216089249,
|
1645 |
+
"learning_rate": 4.44584720007618e-05,
|
1646 |
+
"loss": 0.0197,
|
1647 |
+
"step": 2340
|
1648 |
+
},
|
1649 |
+
{
|
1650 |
+
"epoch": 14.071856287425149,
|
1651 |
+
"grad_norm": 0.13059602677822113,
|
1652 |
+
"learning_rate": 4.363476108191552e-05,
|
1653 |
+
"loss": 0.0144,
|
1654 |
+
"step": 2350
|
1655 |
+
},
|
1656 |
+
{
|
1657 |
+
"epoch": 14.131736526946108,
|
1658 |
+
"grad_norm": 0.1200447678565979,
|
1659 |
+
"learning_rate": 4.2816617641344934e-05,
|
1660 |
+
"loss": 0.0158,
|
1661 |
+
"step": 2360
|
1662 |
+
},
|
1663 |
+
{
|
1664 |
+
"epoch": 14.191616766467066,
|
1665 |
+
"grad_norm": 0.15896056592464447,
|
1666 |
+
"learning_rate": 4.200412249119463e-05,
|
1667 |
+
"loss": 0.0135,
|
1668 |
+
"step": 2370
|
1669 |
+
},
|
1670 |
+
{
|
1671 |
+
"epoch": 14.251497005988025,
|
1672 |
+
"grad_norm": 0.18628185987472534,
|
1673 |
+
"learning_rate": 4.119735588569899e-05,
|
1674 |
+
"loss": 0.0131,
|
1675 |
+
"step": 2380
|
1676 |
+
},
|
1677 |
+
{
|
1678 |
+
"epoch": 14.311377245508982,
|
1679 |
+
"grad_norm": 0.1670549213886261,
|
1680 |
+
"learning_rate": 4.039639751325525e-05,
|
1681 |
+
"loss": 0.0145,
|
1682 |
+
"step": 2390
|
1683 |
+
},
|
1684 |
+
{
|
1685 |
+
"epoch": 14.37125748502994,
|
1686 |
+
"grad_norm": 0.2762540280818939,
|
1687 |
+
"learning_rate": 3.960132648855226e-05,
|
1688 |
+
"loss": 0.0167,
|
1689 |
+
"step": 2400
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 14.431137724550899,
|
1693 |
+
"grad_norm": 0.09425695985555649,
|
1694 |
+
"learning_rate": 3.881222134475589e-05,
|
1695 |
+
"loss": 0.0133,
|
1696 |
+
"step": 2410
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"epoch": 14.491017964071856,
|
1700 |
+
"grad_norm": 0.176064133644104,
|
1701 |
+
"learning_rate": 3.802916002575192e-05,
|
1702 |
+
"loss": 0.0124,
|
1703 |
+
"step": 2420
|
1704 |
+
},
|
1705 |
+
{
|
1706 |
+
"epoch": 14.550898203592814,
|
1707 |
+
"grad_norm": 0.19551633298397064,
|
1708 |
+
"learning_rate": 3.725221987844721e-05,
|
1709 |
+
"loss": 0.0177,
|
1710 |
+
"step": 2430
|
1711 |
+
},
|
1712 |
+
{
|
1713 |
+
"epoch": 14.610778443113773,
|
1714 |
+
"grad_norm": 0.20848046243190765,
|
1715 |
+
"learning_rate": 3.6481477645129624e-05,
|
1716 |
+
"loss": 0.0154,
|
1717 |
+
"step": 2440
|
1718 |
+
},
|
1719 |
+
{
|
1720 |
+
"epoch": 14.67065868263473,
|
1721 |
+
"grad_norm": 0.2565270960330963,
|
1722 |
+
"learning_rate": 3.571700945588789e-05,
|
1723 |
+
"loss": 0.015,
|
1724 |
+
"step": 2450
|
1725 |
+
},
|
1726 |
+
{
|
1727 |
+
"epoch": 14.730538922155688,
|
1728 |
+
"grad_norm": 0.13702988624572754,
|
1729 |
+
"learning_rate": 3.495889082109185e-05,
|
1730 |
+
"loss": 0.0157,
|
1731 |
+
"step": 2460
|
1732 |
+
},
|
1733 |
+
{
|
1734 |
+
"epoch": 14.790419161676647,
|
1735 |
+
"grad_norm": 0.5474858283996582,
|
1736 |
+
"learning_rate": 3.4207196623933857e-05,
|
1737 |
+
"loss": 0.0143,
|
1738 |
+
"step": 2470
|
1739 |
+
},
|
1740 |
+
{
|
1741 |
+
"epoch": 14.850299401197605,
|
1742 |
+
"grad_norm": 0.1655379682779312,
|
1743 |
+
"learning_rate": 3.346200111303226e-05,
|
1744 |
+
"loss": 0.0152,
|
1745 |
+
"step": 2480
|
1746 |
+
},
|
1747 |
+
{
|
1748 |
+
"epoch": 14.910179640718562,
|
1749 |
+
"grad_norm": 0.13053831458091736,
|
1750 |
+
"learning_rate": 3.272337789509751e-05,
|
1751 |
+
"loss": 0.0139,
|
1752 |
+
"step": 2490
|
1753 |
+
},
|
1754 |
+
{
|
1755 |
+
"epoch": 14.970059880239521,
|
1756 |
+
"grad_norm": 0.1283467561006546,
|
1757 |
+
"learning_rate": 3.199139992766155e-05,
|
1758 |
+
"loss": 0.0161,
|
1759 |
+
"step": 2500
|
1760 |
+
},
|
1761 |
+
{
|
1762 |
+
"epoch": 15.029940119760479,
|
1763 |
+
"grad_norm": 0.13677707314491272,
|
1764 |
+
"learning_rate": 3.1266139511871465e-05,
|
1765 |
+
"loss": 0.0126,
|
1766 |
+
"step": 2510
|
1767 |
+
},
|
1768 |
+
{
|
1769 |
+
"epoch": 15.089820359281438,
|
1770 |
+
"grad_norm": 0.10139447450637817,
|
1771 |
+
"learning_rate": 3.0547668285348105e-05,
|
1772 |
+
"loss": 0.0143,
|
1773 |
+
"step": 2520
|
1774 |
+
},
|
1775 |
+
{
|
1776 |
+
"epoch": 15.149700598802395,
|
1777 |
+
"grad_norm": 0.14132164418697357,
|
1778 |
+
"learning_rate": 2.9836057215109746e-05,
|
1779 |
+
"loss": 0.0149,
|
1780 |
+
"step": 2530
|
1781 |
+
},
|
1782 |
+
{
|
1783 |
+
"epoch": 15.209580838323353,
|
1784 |
+
"grad_norm": 0.13676463067531586,
|
1785 |
+
"learning_rate": 2.9131376590562587e-05,
|
1786 |
+
"loss": 0.0131,
|
1787 |
+
"step": 2540
|
1788 |
+
},
|
1789 |
+
{
|
1790 |
+
"epoch": 15.269461077844312,
|
1791 |
+
"grad_norm": 0.20659254491329193,
|
1792 |
+
"learning_rate": 2.843369601655792e-05,
|
1793 |
+
"loss": 0.0125,
|
1794 |
+
"step": 2550
|
1795 |
+
},
|
1796 |
+
{
|
1797 |
+
"epoch": 15.32934131736527,
|
1798 |
+
"grad_norm": 0.13563022017478943,
|
1799 |
+
"learning_rate": 2.7743084406516693e-05,
|
1800 |
+
"loss": 0.0147,
|
1801 |
+
"step": 2560
|
1802 |
+
},
|
1803 |
+
{
|
1804 |
+
"epoch": 15.389221556886227,
|
1805 |
+
"grad_norm": 0.13761834800243378,
|
1806 |
+
"learning_rate": 2.705960997562277e-05,
|
1807 |
+
"loss": 0.0159,
|
1808 |
+
"step": 2570
|
1809 |
+
},
|
1810 |
+
{
|
1811 |
+
"epoch": 15.449101796407186,
|
1812 |
+
"grad_norm": 0.11985321342945099,
|
1813 |
+
"learning_rate": 2.6383340234084875e-05,
|
1814 |
+
"loss": 0.0111,
|
1815 |
+
"step": 2580
|
1816 |
+
},
|
1817 |
+
{
|
1818 |
+
"epoch": 15.508982035928144,
|
1819 |
+
"grad_norm": 0.09939990192651749,
|
1820 |
+
"learning_rate": 2.5714341980468437e-05,
|
1821 |
+
"loss": 0.0122,
|
1822 |
+
"step": 2590
|
1823 |
+
},
|
1824 |
+
{
|
1825 |
+
"epoch": 15.568862275449101,
|
1826 |
+
"grad_norm": 0.19175921380519867,
|
1827 |
+
"learning_rate": 2.5052681295097225e-05,
|
1828 |
+
"loss": 0.0134,
|
1829 |
+
"step": 2600
|
1830 |
+
},
|
1831 |
+
{
|
1832 |
+
"epoch": 15.62874251497006,
|
1833 |
+
"grad_norm": 0.11813875287771225,
|
1834 |
+
"learning_rate": 2.439842353352667e-05,
|
1835 |
+
"loss": 0.015,
|
1836 |
+
"step": 2610
|
1837 |
+
},
|
1838 |
+
{
|
1839 |
+
"epoch": 15.688622754491018,
|
1840 |
+
"grad_norm": 0.16614750027656555,
|
1841 |
+
"learning_rate": 2.3751633320088064e-05,
|
1842 |
+
"loss": 0.0147,
|
1843 |
+
"step": 2620
|
1844 |
+
},
|
1845 |
+
{
|
1846 |
+
"epoch": 15.748502994011975,
|
1847 |
+
"grad_norm": 0.2053995579481125,
|
1848 |
+
"learning_rate": 2.311237454150539e-05,
|
1849 |
+
"loss": 0.0122,
|
1850 |
+
"step": 2630
|
1851 |
+
},
|
1852 |
+
{
|
1853 |
+
"epoch": 15.808383233532934,
|
1854 |
+
"grad_norm": 0.21074947714805603,
|
1855 |
+
"learning_rate": 2.248071034058492e-05,
|
1856 |
+
"loss": 0.0133,
|
1857 |
+
"step": 2640
|
1858 |
+
},
|
1859 |
+
{
|
1860 |
+
"epoch": 15.868263473053892,
|
1861 |
+
"grad_norm": 0.18404161930084229,
|
1862 |
+
"learning_rate": 2.185670310997835e-05,
|
1863 |
+
"loss": 0.0155,
|
1864 |
+
"step": 2650
|
1865 |
+
},
|
1866 |
+
{
|
1867 |
+
"epoch": 15.928143712574851,
|
1868 |
+
"grad_norm": 0.12167193740606308,
|
1869 |
+
"learning_rate": 2.124041448601979e-05,
|
1870 |
+
"loss": 0.0113,
|
1871 |
+
"step": 2660
|
1872 |
+
},
|
1873 |
+
{
|
1874 |
+
"epoch": 15.988023952095809,
|
1875 |
+
"grad_norm": 0.19261722266674042,
|
1876 |
+
"learning_rate": 2.0631905342637782e-05,
|
1877 |
+
"loss": 0.0124,
|
1878 |
+
"step": 2670
|
1879 |
+
},
|
1880 |
+
{
|
1881 |
+
"epoch": 16.047904191616766,
|
1882 |
+
"grad_norm": 0.152290940284729,
|
1883 |
+
"learning_rate": 2.0031235785342407e-05,
|
1884 |
+
"loss": 0.014,
|
1885 |
+
"step": 2680
|
1886 |
+
},
|
1887 |
+
{
|
1888 |
+
"epoch": 16.107784431137723,
|
1889 |
+
"grad_norm": 0.13786451518535614,
|
1890 |
+
"learning_rate": 1.9438465145288374e-05,
|
1891 |
+
"loss": 0.0125,
|
1892 |
+
"step": 2690
|
1893 |
+
},
|
1894 |
+
{
|
1895 |
+
"epoch": 16.16766467065868,
|
1896 |
+
"grad_norm": 0.11830327659845352,
|
1897 |
+
"learning_rate": 1.885365197341461e-05,
|
1898 |
+
"loss": 0.013,
|
1899 |
+
"step": 2700
|
1900 |
+
},
|
1901 |
+
{
|
1902 |
+
"epoch": 16.227544910179642,
|
1903 |
+
"grad_norm": 0.1224045529961586,
|
1904 |
+
"learning_rate": 1.827685403466087e-05,
|
1905 |
+
"loss": 0.0105,
|
1906 |
+
"step": 2710
|
1907 |
+
},
|
1908 |
+
{
|
1909 |
+
"epoch": 16.2874251497006,
|
1910 |
+
"grad_norm": 0.12195927649736404,
|
1911 |
+
"learning_rate": 1.7708128302261994e-05,
|
1912 |
+
"loss": 0.0117,
|
1913 |
+
"step": 2720
|
1914 |
+
},
|
1915 |
+
{
|
1916 |
+
"epoch": 16.347305389221557,
|
1917 |
+
"grad_norm": 0.1412719041109085,
|
1918 |
+
"learning_rate": 1.7147530952120337e-05,
|
1919 |
+
"loss": 0.0131,
|
1920 |
+
"step": 2730
|
1921 |
+
},
|
1922 |
+
{
|
1923 |
+
"epoch": 16.407185628742514,
|
1924 |
+
"grad_norm": 0.1535654515028,
|
1925 |
+
"learning_rate": 1.659511735725716e-05,
|
1926 |
+
"loss": 0.013,
|
1927 |
+
"step": 2740
|
1928 |
+
},
|
1929 |
+
{
|
1930 |
+
"epoch": 16.46706586826347,
|
1931 |
+
"grad_norm": 0.3097449839115143,
|
1932 |
+
"learning_rate": 1.605094208234288e-05,
|
1933 |
+
"loss": 0.0105,
|
1934 |
+
"step": 2750
|
1935 |
+
},
|
1936 |
+
{
|
1937 |
+
"epoch": 16.526946107784433,
|
1938 |
+
"grad_norm": 0.1134740337729454,
|
1939 |
+
"learning_rate": 1.5515058878307665e-05,
|
1940 |
+
"loss": 0.0107,
|
1941 |
+
"step": 2760
|
1942 |
+
},
|
1943 |
+
{
|
1944 |
+
"epoch": 16.58682634730539,
|
1945 |
+
"grad_norm": 0.23593778908252716,
|
1946 |
+
"learning_rate": 1.4987520677032174e-05,
|
1947 |
+
"loss": 0.0117,
|
1948 |
+
"step": 2770
|
1949 |
+
},
|
1950 |
+
{
|
1951 |
+
"epoch": 16.646706586826348,
|
1952 |
+
"grad_norm": 0.1333216279745102,
|
1953 |
+
"learning_rate": 1.4468379586119074e-05,
|
1954 |
+
"loss": 0.0151,
|
1955 |
+
"step": 2780
|
1956 |
+
},
|
1957 |
+
{
|
1958 |
+
"epoch": 16.706586826347305,
|
1959 |
+
"grad_norm": 0.2893410623073578,
|
1960 |
+
"learning_rate": 1.3957686883746191e-05,
|
1961 |
+
"loss": 0.0101,
|
1962 |
+
"step": 2790
|
1963 |
+
},
|
1964 |
+
{
|
1965 |
+
"epoch": 16.766467065868262,
|
1966 |
+
"grad_norm": 0.12125154584646225,
|
1967 |
+
"learning_rate": 1.3455493013601561e-05,
|
1968 |
+
"loss": 0.0115,
|
1969 |
+
"step": 2800
|
1970 |
+
},
|
1971 |
+
{
|
1972 |
+
"epoch": 16.82634730538922,
|
1973 |
+
"grad_norm": 0.3616254925727844,
|
1974 |
+
"learning_rate": 1.2961847579900777e-05,
|
1975 |
+
"loss": 0.0133,
|
1976 |
+
"step": 2810
|
1977 |
+
},
|
1978 |
+
{
|
1979 |
+
"epoch": 16.88622754491018,
|
1980 |
+
"grad_norm": 0.1681334674358368,
|
1981 |
+
"learning_rate": 1.2476799342487255e-05,
|
1982 |
+
"loss": 0.0126,
|
1983 |
+
"step": 2820
|
1984 |
+
},
|
1985 |
+
{
|
1986 |
+
"epoch": 16.94610778443114,
|
1987 |
+
"grad_norm": 0.14123430848121643,
|
1988 |
+
"learning_rate": 1.2000396212016218e-05,
|
1989 |
+
"loss": 0.0146,
|
1990 |
+
"step": 2830
|
1991 |
+
},
|
1992 |
+
{
|
1993 |
+
"epoch": 17.005988023952096,
|
1994 |
+
"grad_norm": 0.15791960060596466,
|
1995 |
+
"learning_rate": 1.1532685245222075e-05,
|
1996 |
+
"loss": 0.0113,
|
1997 |
+
"step": 2840
|
1998 |
+
},
|
1999 |
+
{
|
2000 |
+
"epoch": 17.065868263473053,
|
2001 |
+
"grad_norm": 0.20897255837917328,
|
2002 |
+
"learning_rate": 1.1073712640270506e-05,
|
2003 |
+
"loss": 0.0111,
|
2004 |
+
"step": 2850
|
2005 |
+
},
|
2006 |
+
{
|
2007 |
+
"epoch": 17.12574850299401,
|
2008 |
+
"grad_norm": 0.07108011096715927,
|
2009 |
+
"learning_rate": 1.0623523732195284e-05,
|
2010 |
+
"loss": 0.0123,
|
2011 |
+
"step": 2860
|
2012 |
+
},
|
2013 |
+
{
|
2014 |
+
"epoch": 17.18562874251497,
|
2015 |
+
"grad_norm": 0.2679906189441681,
|
2016 |
+
"learning_rate": 1.0182162988420163e-05,
|
2017 |
+
"loss": 0.0145,
|
2018 |
+
"step": 2870
|
2019 |
+
},
|
2020 |
+
{
|
2021 |
+
"epoch": 17.24550898203593,
|
2022 |
+
"grad_norm": 0.11328988522291183,
|
2023 |
+
"learning_rate": 9.749674004366727e-06,
|
2024 |
+
"loss": 0.0122,
|
2025 |
+
"step": 2880
|
2026 |
+
},
|
2027 |
+
{
|
2028 |
+
"epoch": 17.305389221556887,
|
2029 |
+
"grad_norm": 0.3787294924259186,
|
2030 |
+
"learning_rate": 9.326099499148177e-06,
|
2031 |
+
"loss": 0.0103,
|
2032 |
+
"step": 2890
|
2033 |
+
},
|
2034 |
+
{
|
2035 |
+
"epoch": 17.365269461077844,
|
2036 |
+
"grad_norm": 0.2220940738916397,
|
2037 |
+
"learning_rate": 8.911481311349834e-06,
|
2038 |
+
"loss": 0.0123,
|
2039 |
+
"step": 2900
|
2040 |
+
},
|
2041 |
+
{
|
2042 |
+
"epoch": 17.4251497005988,
|
2043 |
+
"grad_norm": 0.12199822813272476,
|
2044 |
+
"learning_rate": 8.505860394896414e-06,
|
2045 |
+
"loss": 0.011,
|
2046 |
+
"step": 2910
|
2047 |
+
},
|
2048 |
+
{
|
2049 |
+
"epoch": 17.48502994011976,
|
2050 |
+
"grad_norm": 0.10359202325344086,
|
2051 |
+
"learning_rate": 8.109276815006938e-06,
|
2052 |
+
"loss": 0.0102,
|
2053 |
+
"step": 2920
|
2054 |
+
},
|
2055 |
+
{
|
2056 |
+
"epoch": 17.54491017964072,
|
2057 |
+
"grad_norm": 0.49252983927726746,
|
2058 |
+
"learning_rate": 7.721769744237184e-06,
|
2059 |
+
"loss": 0.014,
|
2060 |
+
"step": 2930
|
2061 |
+
},
|
2062 |
+
{
|
2063 |
+
"epoch": 17.604790419161677,
|
2064 |
+
"grad_norm": 0.3275662064552307,
|
2065 |
+
"learning_rate": 7.343377458610446e-06,
|
2066 |
+
"loss": 0.0148,
|
2067 |
+
"step": 2940
|
2068 |
+
},
|
2069 |
+
{
|
2070 |
+
"epoch": 17.664670658682635,
|
2071 |
+
"grad_norm": 0.7134996056556702,
|
2072 |
+
"learning_rate": 6.9741373338368165e-06,
|
2073 |
+
"loss": 0.0129,
|
2074 |
+
"step": 2950
|
2075 |
+
},
|
2076 |
+
{
|
2077 |
+
"epoch": 17.724550898203592,
|
2078 |
+
"grad_norm": 0.10740635544061661,
|
2079 |
+
"learning_rate": 6.614085841621442e-06,
|
2080 |
+
"loss": 0.0111,
|
2081 |
+
"step": 2960
|
2082 |
+
},
|
2083 |
+
{
|
2084 |
+
"epoch": 17.78443113772455,
|
2085 |
+
"grad_norm": 0.13902905583381653,
|
2086 |
+
"learning_rate": 6.263258546061978e-06,
|
2087 |
+
"loss": 0.0124,
|
2088 |
+
"step": 2970
|
2089 |
+
},
|
2090 |
+
{
|
2091 |
+
"epoch": 17.84431137724551,
|
2092 |
+
"grad_norm": 0.2780143916606903,
|
2093 |
+
"learning_rate": 5.921690100135713e-06,
|
2094 |
+
"loss": 0.009,
|
2095 |
+
"step": 2980
|
2096 |
+
},
|
2097 |
+
{
|
2098 |
+
"epoch": 17.904191616766468,
|
2099 |
+
"grad_norm": 0.11749708652496338,
|
2100 |
+
"learning_rate": 5.589414242276869e-06,
|
2101 |
+
"loss": 0.0098,
|
2102 |
+
"step": 2990
|
2103 |
+
},
|
2104 |
+
{
|
2105 |
+
"epoch": 17.964071856287426,
|
2106 |
+
"grad_norm": 0.45455411076545715,
|
2107 |
+
"learning_rate": 5.266463793043896e-06,
|
2108 |
+
"loss": 0.0087,
|
2109 |
+
"step": 3000
|
2110 |
+
},
|
2111 |
+
{
|
2112 |
+
"epoch": 18.023952095808383,
|
2113 |
+
"grad_norm": 0.22297538816928864,
|
2114 |
+
"learning_rate": 4.952870651877739e-06,
|
2115 |
+
"loss": 0.0114,
|
2116 |
+
"step": 3010
|
2117 |
+
},
|
2118 |
+
{
|
2119 |
+
"epoch": 18.08383233532934,
|
2120 |
+
"grad_norm": 0.5311406254768372,
|
2121 |
+
"learning_rate": 4.648665793950968e-06,
|
2122 |
+
"loss": 0.0093,
|
2123 |
+
"step": 3020
|
2124 |
+
},
|
2125 |
+
{
|
2126 |
+
"epoch": 18.143712574850298,
|
2127 |
+
"grad_norm": 0.20427249372005463,
|
2128 |
+
"learning_rate": 4.353879267108119e-06,
|
2129 |
+
"loss": 0.0123,
|
2130 |
+
"step": 3030
|
2131 |
+
},
|
2132 |
+
{
|
2133 |
+
"epoch": 18.20359281437126,
|
2134 |
+
"grad_norm": 0.2326945662498474,
|
2135 |
+
"learning_rate": 4.068540188897796e-06,
|
2136 |
+
"loss": 0.0099,
|
2137 |
+
"step": 3040
|
2138 |
+
},
|
2139 |
+
{
|
2140 |
+
"epoch": 18.263473053892216,
|
2141 |
+
"grad_norm": 0.14544923603534698,
|
2142 |
+
"learning_rate": 3.792676743696588e-06,
|
2143 |
+
"loss": 0.0123,
|
2144 |
+
"step": 3050
|
2145 |
+
},
|
2146 |
+
{
|
2147 |
+
"epoch": 18.323353293413174,
|
2148 |
+
"grad_norm": 0.1755136400461197,
|
2149 |
+
"learning_rate": 3.5263161799251154e-06,
|
2150 |
+
"loss": 0.0095,
|
2151 |
+
"step": 3060
|
2152 |
+
},
|
2153 |
+
{
|
2154 |
+
"epoch": 18.38323353293413,
|
2155 |
+
"grad_norm": 0.10239781439304352,
|
2156 |
+
"learning_rate": 3.2694848073565596e-06,
|
2157 |
+
"loss": 0.0095,
|
2158 |
+
"step": 3070
|
2159 |
+
},
|
2160 |
+
{
|
2161 |
+
"epoch": 18.44311377245509,
|
2162 |
+
"grad_norm": 0.13217875361442566,
|
2163 |
+
"learning_rate": 3.0222079945179895e-06,
|
2164 |
+
"loss": 0.0111,
|
2165 |
+
"step": 3080
|
2166 |
+
},
|
2167 |
+
{
|
2168 |
+
"epoch": 18.50299401197605,
|
2169 |
+
"grad_norm": 0.11932236701250076,
|
2170 |
+
"learning_rate": 2.7845101661844618e-06,
|
2171 |
+
"loss": 0.0099,
|
2172 |
+
"step": 3090
|
2173 |
+
},
|
2174 |
+
{
|
2175 |
+
"epoch": 18.562874251497007,
|
2176 |
+
"grad_norm": 0.12827470898628235,
|
2177 |
+
"learning_rate": 2.5564148009665778e-06,
|
2178 |
+
"loss": 0.0125,
|
2179 |
+
"step": 3100
|
2180 |
+
},
|
2181 |
+
{
|
2182 |
+
"epoch": 18.622754491017965,
|
2183 |
+
"grad_norm": 0.44992345571517944,
|
2184 |
+
"learning_rate": 2.3379444289913342e-06,
|
2185 |
+
"loss": 0.0114,
|
2186 |
+
"step": 3110
|
2187 |
+
},
|
2188 |
+
{
|
2189 |
+
"epoch": 18.682634730538922,
|
2190 |
+
"grad_norm": 0.22988931834697723,
|
2191 |
+
"learning_rate": 2.1291206296767063e-06,
|
2192 |
+
"loss": 0.0121,
|
2193 |
+
"step": 3120
|
2194 |
+
},
|
2195 |
+
{
|
2196 |
+
"epoch": 18.74251497005988,
|
2197 |
+
"grad_norm": 0.2384101152420044,
|
2198 |
+
"learning_rate": 1.9299640296001288e-06,
|
2199 |
+
"loss": 0.0094,
|
2200 |
+
"step": 3130
|
2201 |
+
},
|
2202 |
+
{
|
2203 |
+
"epoch": 18.802395209580837,
|
2204 |
+
"grad_norm": 0.10543084144592285,
|
2205 |
+
"learning_rate": 1.7404943004611596e-06,
|
2206 |
+
"loss": 0.013,
|
2207 |
+
"step": 3140
|
2208 |
+
},
|
2209 |
+
{
|
2210 |
+
"epoch": 18.862275449101798,
|
2211 |
+
"grad_norm": 0.12687674164772034,
|
2212 |
+
"learning_rate": 1.5607301571383459e-06,
|
2213 |
+
"loss": 0.0112,
|
2214 |
+
"step": 3150
|
2215 |
+
},
|
2216 |
+
{
|
2217 |
+
"epoch": 18.922155688622755,
|
2218 |
+
"grad_norm": 0.26404881477355957,
|
2219 |
+
"learning_rate": 1.390689355840691e-06,
|
2220 |
+
"loss": 0.0133,
|
2221 |
+
"step": 3160
|
2222 |
+
},
|
2223 |
+
{
|
2224 |
+
"epoch": 18.982035928143713,
|
2225 |
+
"grad_norm": 0.11557133495807648,
|
2226 |
+
"learning_rate": 1.2303886923537677e-06,
|
2227 |
+
"loss": 0.0107,
|
2228 |
+
"step": 3170
|
2229 |
+
},
|
2230 |
+
{
|
2231 |
+
"epoch": 19.04191616766467,
|
2232 |
+
"grad_norm": 0.10522342473268509,
|
2233 |
+
"learning_rate": 1.0798440003807474e-06,
|
2234 |
+
"loss": 0.0107,
|
2235 |
+
"step": 3180
|
2236 |
+
},
|
2237 |
+
{
|
2238 |
+
"epoch": 19.101796407185628,
|
2239 |
+
"grad_norm": 0.1134905070066452,
|
2240 |
+
"learning_rate": 9.390701499783827e-07,
|
2241 |
+
"loss": 0.0086,
|
2242 |
+
"step": 3190
|
2243 |
+
},
|
2244 |
+
{
|
2245 |
+
"epoch": 19.161676646706585,
|
2246 |
+
"grad_norm": 0.1318759173154831,
|
2247 |
+
"learning_rate": 8.080810460882493e-07,
|
2248 |
+
"loss": 0.0107,
|
2249 |
+
"step": 3200
|
2250 |
+
},
|
2251 |
+
{
|
2252 |
+
"epoch": 19.221556886227546,
|
2253 |
+
"grad_norm": 0.11792134493589401,
|
2254 |
+
"learning_rate": 6.868896271632785e-07,
|
2255 |
+
"loss": 0.0099,
|
2256 |
+
"step": 3210
|
2257 |
+
},
|
2258 |
+
{
|
2259 |
+
"epoch": 19.281437125748504,
|
2260 |
+
"grad_norm": 0.18841320276260376,
|
2261 |
+
"learning_rate": 5.755078638897459e-07,
|
2262 |
+
"loss": 0.0118,
|
2263 |
+
"step": 3220
|
2264 |
+
},
|
2265 |
+
{
|
2266 |
+
"epoch": 19.34131736526946,
|
2267 |
+
"grad_norm": 0.1288711428642273,
|
2268 |
+
"learning_rate": 4.739467580048618e-07,
|
2269 |
+
"loss": 0.0092,
|
2270 |
+
"step": 3230
|
2271 |
+
},
|
2272 |
+
{
|
2273 |
+
"epoch": 19.40119760479042,
|
2274 |
+
"grad_norm": 0.3439689874649048,
|
2275 |
+
"learning_rate": 3.8221634121013983e-07,
|
2276 |
+
"loss": 0.0083,
|
2277 |
+
"step": 3240
|
2278 |
+
},
|
2279 |
+
{
|
2280 |
+
"epoch": 19.461077844311376,
|
2281 |
+
"grad_norm": 0.07992304116487503,
|
2282 |
+
"learning_rate": 3.003256741803906e-07,
|
2283 |
+
"loss": 0.012,
|
2284 |
+
"step": 3250
|
2285 |
+
},
|
2286 |
+
{
|
2287 |
+
"epoch": 19.520958083832337,
|
2288 |
+
"grad_norm": 0.15533998608589172,
|
2289 |
+
"learning_rate": 2.2828284566890345e-07,
|
2290 |
+
"loss": 0.0108,
|
2291 |
+
"step": 3260
|
2292 |
+
},
|
2293 |
+
{
|
2294 |
+
"epoch": 19.580838323353294,
|
2295 |
+
"grad_norm": 0.3759068548679352,
|
2296 |
+
"learning_rate": 1.6609497170834155e-07,
|
2297 |
+
"loss": 0.0112,
|
2298 |
+
"step": 3270
|
2299 |
+
},
|
2300 |
+
{
|
2301 |
+
"epoch": 19.64071856287425,
|
2302 |
+
"grad_norm": 0.12365289032459259,
|
2303 |
+
"learning_rate": 1.1376819490795942e-07,
|
2304 |
+
"loss": 0.0135,
|
2305 |
+
"step": 3280
|
2306 |
+
},
|
2307 |
+
{
|
2308 |
+
"epoch": 19.70059880239521,
|
2309 |
+
"grad_norm": 0.12051455676555634,
|
2310 |
+
"learning_rate": 7.130768384679965e-08,
|
2311 |
+
"loss": 0.01,
|
2312 |
+
"step": 3290
|
2313 |
+
},
|
2314 |
+
{
|
2315 |
+
"epoch": 19.760479041916167,
|
2316 |
+
"grad_norm": 0.10926169902086258,
|
2317 |
+
"learning_rate": 3.8717632563201046e-08,
|
2318 |
+
"loss": 0.0152,
|
2319 |
+
"step": 3300
|
2320 |
+
},
|
2321 |
+
{
|
2322 |
+
"epoch": 19.820359281437124,
|
2323 |
+
"grad_norm": 0.15941037237644196,
|
2324 |
+
"learning_rate": 1.600126014051906e-08,
|
2325 |
+
"loss": 0.0109,
|
2326 |
+
"step": 3310
|
2327 |
+
},
|
2328 |
+
{
|
2329 |
+
"epoch": 19.880239520958085,
|
2330 |
+
"grad_norm": 0.07696834206581116,
|
2331 |
+
"learning_rate": 3.1608103891689335e-09,
|
2332 |
+
"loss": 0.0093,
|
2333 |
+
"step": 3320
|
2334 |
+
},
|
2335 |
+
{
|
2336 |
+
"epoch": 19.92814371257485,
|
2337 |
+
"step": 3328,
|
2338 |
+
"total_flos": 4.678773863650698e+17,
|
2339 |
+
"train_loss": 0.034036972624912426,
|
2340 |
+
"train_runtime": 3742.5882,
|
2341 |
+
"train_samples_per_second": 56.91,
|
2342 |
+
"train_steps_per_second": 0.889
|
2343 |
+
}
|
2344 |
+
],
|
2345 |
+
"logging_steps": 10,
|
2346 |
+
"max_steps": 3328,
|
2347 |
+
"num_input_tokens_seen": 0,
|
2348 |
+
"num_train_epochs": 20,
|
2349 |
+
"save_steps": 10000,
|
2350 |
+
"stateful_callbacks": {
|
2351 |
+
"TrainerControl": {
|
2352 |
+
"args": {
|
2353 |
+
"should_epoch_stop": false,
|
2354 |
+
"should_evaluate": false,
|
2355 |
+
"should_log": false,
|
2356 |
+
"should_save": true,
|
2357 |
+
"should_training_stop": true
|
2358 |
+
},
|
2359 |
+
"attributes": {}
|
2360 |
+
}
|
2361 |
+
},
|
2362 |
+
"total_flos": 4.678773863650698e+17,
|
2363 |
+
"train_batch_size": 64,
|
2364 |
+
"trial_name": null,
|
2365 |
+
"trial_params": null
|
2366 |
+
}
|