Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +3473 -0
trainer_state.json
ADDED
@@ -0,0 +1,3473 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_global_step": null,
|
3 |
+
"best_metric": null,
|
4 |
+
"best_model_checkpoint": null,
|
5 |
+
"epoch": 10.0,
|
6 |
+
"eval_steps": 500,
|
7 |
+
"global_step": 4900,
|
8 |
+
"is_hyper_param_search": false,
|
9 |
+
"is_local_process_zero": true,
|
10 |
+
"is_world_process_zero": true,
|
11 |
+
"log_history": [
|
12 |
+
{
|
13 |
+
"epoch": 0.02040816326530612,
|
14 |
+
"grad_norm": 10.628430366516113,
|
15 |
+
"learning_rate": 3.6734693877551024e-06,
|
16 |
+
"loss": 1.0152,
|
17 |
+
"step": 10
|
18 |
+
},
|
19 |
+
{
|
20 |
+
"epoch": 0.04081632653061224,
|
21 |
+
"grad_norm": 3.175971269607544,
|
22 |
+
"learning_rate": 7.755102040816327e-06,
|
23 |
+
"loss": 0.5298,
|
24 |
+
"step": 20
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"epoch": 0.061224489795918366,
|
28 |
+
"grad_norm": 2.9172680377960205,
|
29 |
+
"learning_rate": 1.1836734693877552e-05,
|
30 |
+
"loss": 0.337,
|
31 |
+
"step": 30
|
32 |
+
},
|
33 |
+
{
|
34 |
+
"epoch": 0.08163265306122448,
|
35 |
+
"grad_norm": 2.0697555541992188,
|
36 |
+
"learning_rate": 1.5918367346938776e-05,
|
37 |
+
"loss": 0.255,
|
38 |
+
"step": 40
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 0.10204081632653061,
|
42 |
+
"grad_norm": 1.7489535808563232,
|
43 |
+
"learning_rate": 2e-05,
|
44 |
+
"loss": 0.2925,
|
45 |
+
"step": 50
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 0.12244897959183673,
|
49 |
+
"grad_norm": 2.501734495162964,
|
50 |
+
"learning_rate": 2.4081632653061224e-05,
|
51 |
+
"loss": 0.2364,
|
52 |
+
"step": 60
|
53 |
+
},
|
54 |
+
{
|
55 |
+
"epoch": 0.14285714285714285,
|
56 |
+
"grad_norm": 1.3701616525650024,
|
57 |
+
"learning_rate": 2.816326530612245e-05,
|
58 |
+
"loss": 0.2546,
|
59 |
+
"step": 70
|
60 |
+
},
|
61 |
+
{
|
62 |
+
"epoch": 0.16326530612244897,
|
63 |
+
"grad_norm": 2.0604536533355713,
|
64 |
+
"learning_rate": 3.224489795918367e-05,
|
65 |
+
"loss": 0.2513,
|
66 |
+
"step": 80
|
67 |
+
},
|
68 |
+
{
|
69 |
+
"epoch": 0.1836734693877551,
|
70 |
+
"grad_norm": 1.1776974201202393,
|
71 |
+
"learning_rate": 3.63265306122449e-05,
|
72 |
+
"loss": 0.2447,
|
73 |
+
"step": 90
|
74 |
+
},
|
75 |
+
{
|
76 |
+
"epoch": 0.20408163265306123,
|
77 |
+
"grad_norm": 1.3436124324798584,
|
78 |
+
"learning_rate": 4.0408163265306124e-05,
|
79 |
+
"loss": 0.2096,
|
80 |
+
"step": 100
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 0.22448979591836735,
|
84 |
+
"grad_norm": 1.3945685625076294,
|
85 |
+
"learning_rate": 4.448979591836735e-05,
|
86 |
+
"loss": 0.1879,
|
87 |
+
"step": 110
|
88 |
+
},
|
89 |
+
{
|
90 |
+
"epoch": 0.24489795918367346,
|
91 |
+
"grad_norm": 1.264801263809204,
|
92 |
+
"learning_rate": 4.8571428571428576e-05,
|
93 |
+
"loss": 0.1622,
|
94 |
+
"step": 120
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"epoch": 0.2653061224489796,
|
98 |
+
"grad_norm": 1.1072674989700317,
|
99 |
+
"learning_rate": 5.26530612244898e-05,
|
100 |
+
"loss": 0.1611,
|
101 |
+
"step": 130
|
102 |
+
},
|
103 |
+
{
|
104 |
+
"epoch": 0.2857142857142857,
|
105 |
+
"grad_norm": 1.289307713508606,
|
106 |
+
"learning_rate": 5.673469387755103e-05,
|
107 |
+
"loss": 0.133,
|
108 |
+
"step": 140
|
109 |
+
},
|
110 |
+
{
|
111 |
+
"epoch": 0.30612244897959184,
|
112 |
+
"grad_norm": 1.7481720447540283,
|
113 |
+
"learning_rate": 6.081632653061224e-05,
|
114 |
+
"loss": 0.1362,
|
115 |
+
"step": 150
|
116 |
+
},
|
117 |
+
{
|
118 |
+
"epoch": 0.32653061224489793,
|
119 |
+
"grad_norm": 1.594106912612915,
|
120 |
+
"learning_rate": 6.489795918367347e-05,
|
121 |
+
"loss": 0.1247,
|
122 |
+
"step": 160
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 0.3469387755102041,
|
126 |
+
"grad_norm": 1.075558066368103,
|
127 |
+
"learning_rate": 6.897959183673471e-05,
|
128 |
+
"loss": 0.125,
|
129 |
+
"step": 170
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 0.3673469387755102,
|
133 |
+
"grad_norm": 1.093751311302185,
|
134 |
+
"learning_rate": 7.306122448979592e-05,
|
135 |
+
"loss": 0.1133,
|
136 |
+
"step": 180
|
137 |
+
},
|
138 |
+
{
|
139 |
+
"epoch": 0.3877551020408163,
|
140 |
+
"grad_norm": 0.7638540863990784,
|
141 |
+
"learning_rate": 7.714285714285715e-05,
|
142 |
+
"loss": 0.1103,
|
143 |
+
"step": 190
|
144 |
+
},
|
145 |
+
{
|
146 |
+
"epoch": 0.40816326530612246,
|
147 |
+
"grad_norm": 0.8106374740600586,
|
148 |
+
"learning_rate": 8.122448979591836e-05,
|
149 |
+
"loss": 0.1062,
|
150 |
+
"step": 200
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 0.42857142857142855,
|
154 |
+
"grad_norm": 0.8688479065895081,
|
155 |
+
"learning_rate": 8.53061224489796e-05,
|
156 |
+
"loss": 0.1145,
|
157 |
+
"step": 210
|
158 |
+
},
|
159 |
+
{
|
160 |
+
"epoch": 0.4489795918367347,
|
161 |
+
"grad_norm": 1.0158841609954834,
|
162 |
+
"learning_rate": 8.938775510204082e-05,
|
163 |
+
"loss": 0.1099,
|
164 |
+
"step": 220
|
165 |
+
},
|
166 |
+
{
|
167 |
+
"epoch": 0.46938775510204084,
|
168 |
+
"grad_norm": 0.9198583364486694,
|
169 |
+
"learning_rate": 9.346938775510204e-05,
|
170 |
+
"loss": 0.1003,
|
171 |
+
"step": 230
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 0.4897959183673469,
|
175 |
+
"grad_norm": 1.484209418296814,
|
176 |
+
"learning_rate": 9.755102040816328e-05,
|
177 |
+
"loss": 0.105,
|
178 |
+
"step": 240
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 0.5102040816326531,
|
182 |
+
"grad_norm": 0.7159306406974792,
|
183 |
+
"learning_rate": 9.999981781185989e-05,
|
184 |
+
"loss": 0.104,
|
185 |
+
"step": 250
|
186 |
+
},
|
187 |
+
{
|
188 |
+
"epoch": 0.5306122448979592,
|
189 |
+
"grad_norm": 0.8900401592254639,
|
190 |
+
"learning_rate": 9.999776821053133e-05,
|
191 |
+
"loss": 0.0946,
|
192 |
+
"step": 260
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 0.5510204081632653,
|
196 |
+
"grad_norm": 0.9417662620544434,
|
197 |
+
"learning_rate": 9.999344136636323e-05,
|
198 |
+
"loss": 0.1161,
|
199 |
+
"step": 270
|
200 |
+
},
|
201 |
+
{
|
202 |
+
"epoch": 0.5714285714285714,
|
203 |
+
"grad_norm": 0.7933630347251892,
|
204 |
+
"learning_rate": 9.998683747642985e-05,
|
205 |
+
"loss": 0.0966,
|
206 |
+
"step": 280
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 0.5918367346938775,
|
210 |
+
"grad_norm": 1.1042319536209106,
|
211 |
+
"learning_rate": 9.997795684151788e-05,
|
212 |
+
"loss": 0.113,
|
213 |
+
"step": 290
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 0.6122448979591837,
|
217 |
+
"grad_norm": 0.4663528501987457,
|
218 |
+
"learning_rate": 9.996679986611258e-05,
|
219 |
+
"loss": 0.096,
|
220 |
+
"step": 300
|
221 |
+
},
|
222 |
+
{
|
223 |
+
"epoch": 0.6326530612244898,
|
224 |
+
"grad_norm": 1.089492678642273,
|
225 |
+
"learning_rate": 9.99533670583795e-05,
|
226 |
+
"loss": 0.1061,
|
227 |
+
"step": 310
|
228 |
+
},
|
229 |
+
{
|
230 |
+
"epoch": 0.6530612244897959,
|
231 |
+
"grad_norm": 0.8293689489364624,
|
232 |
+
"learning_rate": 9.993765903014125e-05,
|
233 |
+
"loss": 0.0948,
|
234 |
+
"step": 320
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 0.673469387755102,
|
238 |
+
"grad_norm": 0.7057058811187744,
|
239 |
+
"learning_rate": 9.991967649684967e-05,
|
240 |
+
"loss": 0.0966,
|
241 |
+
"step": 330
|
242 |
+
},
|
243 |
+
{
|
244 |
+
"epoch": 0.6938775510204082,
|
245 |
+
"grad_norm": 0.5784198641777039,
|
246 |
+
"learning_rate": 9.98994202775532e-05,
|
247 |
+
"loss": 0.0875,
|
248 |
+
"step": 340
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 0.7142857142857143,
|
252 |
+
"grad_norm": 0.6503363847732544,
|
253 |
+
"learning_rate": 9.987689129485964e-05,
|
254 |
+
"loss": 0.0929,
|
255 |
+
"step": 350
|
256 |
+
},
|
257 |
+
{
|
258 |
+
"epoch": 0.7346938775510204,
|
259 |
+
"grad_norm": 0.7261630892753601,
|
260 |
+
"learning_rate": 9.98520905748941e-05,
|
261 |
+
"loss": 0.0823,
|
262 |
+
"step": 360
|
263 |
+
},
|
264 |
+
{
|
265 |
+
"epoch": 0.7551020408163265,
|
266 |
+
"grad_norm": 0.8498600721359253,
|
267 |
+
"learning_rate": 9.982501924725222e-05,
|
268 |
+
"loss": 0.0918,
|
269 |
+
"step": 370
|
270 |
+
},
|
271 |
+
{
|
272 |
+
"epoch": 0.7755102040816326,
|
273 |
+
"grad_norm": 0.5623912215232849,
|
274 |
+
"learning_rate": 9.979567854494877e-05,
|
275 |
+
"loss": 0.0873,
|
276 |
+
"step": 380
|
277 |
+
},
|
278 |
+
{
|
279 |
+
"epoch": 0.7959183673469388,
|
280 |
+
"grad_norm": 0.5773722529411316,
|
281 |
+
"learning_rate": 9.97640698043615e-05,
|
282 |
+
"loss": 0.0921,
|
283 |
+
"step": 390
|
284 |
+
},
|
285 |
+
{
|
286 |
+
"epoch": 0.8163265306122449,
|
287 |
+
"grad_norm": 0.6975867748260498,
|
288 |
+
"learning_rate": 9.973019446517023e-05,
|
289 |
+
"loss": 0.0821,
|
290 |
+
"step": 400
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 0.8367346938775511,
|
294 |
+
"grad_norm": 0.7360525131225586,
|
295 |
+
"learning_rate": 9.96940540702913e-05,
|
296 |
+
"loss": 0.0965,
|
297 |
+
"step": 410
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 0.8571428571428571,
|
301 |
+
"grad_norm": 1.1155353784561157,
|
302 |
+
"learning_rate": 9.96556502658073e-05,
|
303 |
+
"loss": 0.0879,
|
304 |
+
"step": 420
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 0.8775510204081632,
|
308 |
+
"grad_norm": 0.43246451020240784,
|
309 |
+
"learning_rate": 9.961498480089208e-05,
|
310 |
+
"loss": 0.0892,
|
311 |
+
"step": 430
|
312 |
+
},
|
313 |
+
{
|
314 |
+
"epoch": 0.8979591836734694,
|
315 |
+
"grad_norm": 0.6282567381858826,
|
316 |
+
"learning_rate": 9.957205952773108e-05,
|
317 |
+
"loss": 0.0826,
|
318 |
+
"step": 440
|
319 |
+
},
|
320 |
+
{
|
321 |
+
"epoch": 0.9183673469387755,
|
322 |
+
"grad_norm": 0.9212186932563782,
|
323 |
+
"learning_rate": 9.952687640143699e-05,
|
324 |
+
"loss": 0.0901,
|
325 |
+
"step": 450
|
326 |
+
},
|
327 |
+
{
|
328 |
+
"epoch": 0.9387755102040817,
|
329 |
+
"grad_norm": 0.7539145350456238,
|
330 |
+
"learning_rate": 9.947943747996069e-05,
|
331 |
+
"loss": 0.0881,
|
332 |
+
"step": 460
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 0.9591836734693877,
|
336 |
+
"grad_norm": 0.4982381761074066,
|
337 |
+
"learning_rate": 9.94297449239975e-05,
|
338 |
+
"loss": 0.0895,
|
339 |
+
"step": 470
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 0.9795918367346939,
|
343 |
+
"grad_norm": 0.6672772169113159,
|
344 |
+
"learning_rate": 9.93778009968888e-05,
|
345 |
+
"loss": 0.081,
|
346 |
+
"step": 480
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"epoch": 1.0,
|
350 |
+
"grad_norm": 0.8654345870018005,
|
351 |
+
"learning_rate": 9.932360806451893e-05,
|
352 |
+
"loss": 0.084,
|
353 |
+
"step": 490
|
354 |
+
},
|
355 |
+
{
|
356 |
+
"epoch": 1.0204081632653061,
|
357 |
+
"grad_norm": 0.7786339521408081,
|
358 |
+
"learning_rate": 9.926716859520737e-05,
|
359 |
+
"loss": 0.0846,
|
360 |
+
"step": 500
|
361 |
+
},
|
362 |
+
{
|
363 |
+
"epoch": 1.0408163265306123,
|
364 |
+
"grad_norm": 1.0392732620239258,
|
365 |
+
"learning_rate": 9.920848515959645e-05,
|
366 |
+
"loss": 0.0846,
|
367 |
+
"step": 510
|
368 |
+
},
|
369 |
+
{
|
370 |
+
"epoch": 1.0612244897959184,
|
371 |
+
"grad_norm": 0.7900115847587585,
|
372 |
+
"learning_rate": 9.91475604305341e-05,
|
373 |
+
"loss": 0.0749,
|
374 |
+
"step": 520
|
375 |
+
},
|
376 |
+
{
|
377 |
+
"epoch": 1.0816326530612246,
|
378 |
+
"grad_norm": 0.5556976795196533,
|
379 |
+
"learning_rate": 9.908439718295229e-05,
|
380 |
+
"loss": 0.0785,
|
381 |
+
"step": 530
|
382 |
+
},
|
383 |
+
{
|
384 |
+
"epoch": 1.1020408163265305,
|
385 |
+
"grad_norm": 0.5236167311668396,
|
386 |
+
"learning_rate": 9.901899829374047e-05,
|
387 |
+
"loss": 0.0773,
|
388 |
+
"step": 540
|
389 |
+
},
|
390 |
+
{
|
391 |
+
"epoch": 1.1224489795918366,
|
392 |
+
"grad_norm": 1.0668035745620728,
|
393 |
+
"learning_rate": 9.895136674161465e-05,
|
394 |
+
"loss": 0.0769,
|
395 |
+
"step": 550
|
396 |
+
},
|
397 |
+
{
|
398 |
+
"epoch": 1.1428571428571428,
|
399 |
+
"grad_norm": 0.6634009480476379,
|
400 |
+
"learning_rate": 9.888150560698169e-05,
|
401 |
+
"loss": 0.065,
|
402 |
+
"step": 560
|
403 |
+
},
|
404 |
+
{
|
405 |
+
"epoch": 1.163265306122449,
|
406 |
+
"grad_norm": 0.6299061179161072,
|
407 |
+
"learning_rate": 9.880941807179895e-05,
|
408 |
+
"loss": 0.0791,
|
409 |
+
"step": 570
|
410 |
+
},
|
411 |
+
{
|
412 |
+
"epoch": 1.183673469387755,
|
413 |
+
"grad_norm": 0.8009123802185059,
|
414 |
+
"learning_rate": 9.873510741942951e-05,
|
415 |
+
"loss": 0.0854,
|
416 |
+
"step": 580
|
417 |
+
},
|
418 |
+
{
|
419 |
+
"epoch": 1.2040816326530612,
|
420 |
+
"grad_norm": 0.7957753539085388,
|
421 |
+
"learning_rate": 9.865857703449242e-05,
|
422 |
+
"loss": 0.0826,
|
423 |
+
"step": 590
|
424 |
+
},
|
425 |
+
{
|
426 |
+
"epoch": 1.2244897959183674,
|
427 |
+
"grad_norm": 0.7319406270980835,
|
428 |
+
"learning_rate": 9.857983040270872e-05,
|
429 |
+
"loss": 0.0849,
|
430 |
+
"step": 600
|
431 |
+
},
|
432 |
+
{
|
433 |
+
"epoch": 1.2448979591836735,
|
434 |
+
"grad_norm": 0.7303819060325623,
|
435 |
+
"learning_rate": 9.849887111074256e-05,
|
436 |
+
"loss": 0.0856,
|
437 |
+
"step": 610
|
438 |
+
},
|
439 |
+
{
|
440 |
+
"epoch": 1.2653061224489797,
|
441 |
+
"grad_norm": 0.5338705778121948,
|
442 |
+
"learning_rate": 9.841570284603787e-05,
|
443 |
+
"loss": 0.0742,
|
444 |
+
"step": 620
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"epoch": 1.2857142857142856,
|
448 |
+
"grad_norm": 0.5050308704376221,
|
449 |
+
"learning_rate": 9.833032939665047e-05,
|
450 |
+
"loss": 0.0693,
|
451 |
+
"step": 630
|
452 |
+
},
|
453 |
+
{
|
454 |
+
"epoch": 1.306122448979592,
|
455 |
+
"grad_norm": 0.6604131460189819,
|
456 |
+
"learning_rate": 9.824275465107545e-05,
|
457 |
+
"loss": 0.0695,
|
458 |
+
"step": 640
|
459 |
+
},
|
460 |
+
{
|
461 |
+
"epoch": 1.3265306122448979,
|
462 |
+
"grad_norm": 0.691947877407074,
|
463 |
+
"learning_rate": 9.815298259807008e-05,
|
464 |
+
"loss": 0.0713,
|
465 |
+
"step": 650
|
466 |
+
},
|
467 |
+
{
|
468 |
+
"epoch": 1.346938775510204,
|
469 |
+
"grad_norm": 0.5263407826423645,
|
470 |
+
"learning_rate": 9.806101732647217e-05,
|
471 |
+
"loss": 0.0743,
|
472 |
+
"step": 660
|
473 |
+
},
|
474 |
+
{
|
475 |
+
"epoch": 1.3673469387755102,
|
476 |
+
"grad_norm": 0.8018398880958557,
|
477 |
+
"learning_rate": 9.796686302501382e-05,
|
478 |
+
"loss": 0.0791,
|
479 |
+
"step": 670
|
480 |
+
},
|
481 |
+
{
|
482 |
+
"epoch": 1.3877551020408163,
|
483 |
+
"grad_norm": 0.7871843576431274,
|
484 |
+
"learning_rate": 9.787052398213062e-05,
|
485 |
+
"loss": 0.0792,
|
486 |
+
"step": 680
|
487 |
+
},
|
488 |
+
{
|
489 |
+
"epoch": 1.4081632653061225,
|
490 |
+
"grad_norm": 0.509185254573822,
|
491 |
+
"learning_rate": 9.777200458576634e-05,
|
492 |
+
"loss": 0.0778,
|
493 |
+
"step": 690
|
494 |
+
},
|
495 |
+
{
|
496 |
+
"epoch": 1.4285714285714286,
|
497 |
+
"grad_norm": 0.6663973927497864,
|
498 |
+
"learning_rate": 9.767130932317305e-05,
|
499 |
+
"loss": 0.0769,
|
500 |
+
"step": 700
|
501 |
+
},
|
502 |
+
{
|
503 |
+
"epoch": 1.4489795918367347,
|
504 |
+
"grad_norm": 0.6002593636512756,
|
505 |
+
"learning_rate": 9.756844278070682e-05,
|
506 |
+
"loss": 0.0775,
|
507 |
+
"step": 710
|
508 |
+
},
|
509 |
+
{
|
510 |
+
"epoch": 1.469387755102041,
|
511 |
+
"grad_norm": 0.5615339875221252,
|
512 |
+
"learning_rate": 9.746340964361871e-05,
|
513 |
+
"loss": 0.0672,
|
514 |
+
"step": 720
|
515 |
+
},
|
516 |
+
{
|
517 |
+
"epoch": 1.489795918367347,
|
518 |
+
"grad_norm": 1.005101203918457,
|
519 |
+
"learning_rate": 9.735621469584144e-05,
|
520 |
+
"loss": 0.0762,
|
521 |
+
"step": 730
|
522 |
+
},
|
523 |
+
{
|
524 |
+
"epoch": 1.510204081632653,
|
525 |
+
"grad_norm": 0.7237843871116638,
|
526 |
+
"learning_rate": 9.724686281977146e-05,
|
527 |
+
"loss": 0.0643,
|
528 |
+
"step": 740
|
529 |
+
},
|
530 |
+
{
|
531 |
+
"epoch": 1.5306122448979593,
|
532 |
+
"grad_norm": 0.4733352065086365,
|
533 |
+
"learning_rate": 9.713535899604666e-05,
|
534 |
+
"loss": 0.0717,
|
535 |
+
"step": 750
|
536 |
+
},
|
537 |
+
{
|
538 |
+
"epoch": 1.5510204081632653,
|
539 |
+
"grad_norm": 0.5426393151283264,
|
540 |
+
"learning_rate": 9.70217083033194e-05,
|
541 |
+
"loss": 0.0648,
|
542 |
+
"step": 760
|
543 |
+
},
|
544 |
+
{
|
545 |
+
"epoch": 1.5714285714285714,
|
546 |
+
"grad_norm": 0.7964388132095337,
|
547 |
+
"learning_rate": 9.690591591802524e-05,
|
548 |
+
"loss": 0.0701,
|
549 |
+
"step": 770
|
550 |
+
},
|
551 |
+
{
|
552 |
+
"epoch": 1.5918367346938775,
|
553 |
+
"grad_norm": 0.6255714893341064,
|
554 |
+
"learning_rate": 9.678798711414721e-05,
|
555 |
+
"loss": 0.0694,
|
556 |
+
"step": 780
|
557 |
+
},
|
558 |
+
{
|
559 |
+
"epoch": 1.6122448979591837,
|
560 |
+
"grad_norm": 0.5886476635932922,
|
561 |
+
"learning_rate": 9.666792726297555e-05,
|
562 |
+
"loss": 0.0652,
|
563 |
+
"step": 790
|
564 |
+
},
|
565 |
+
{
|
566 |
+
"epoch": 1.6326530612244898,
|
567 |
+
"grad_norm": 0.5296547412872314,
|
568 |
+
"learning_rate": 9.654574183286307e-05,
|
569 |
+
"loss": 0.0653,
|
570 |
+
"step": 800
|
571 |
+
},
|
572 |
+
{
|
573 |
+
"epoch": 1.6530612244897958,
|
574 |
+
"grad_norm": 0.7035044431686401,
|
575 |
+
"learning_rate": 9.642143638897609e-05,
|
576 |
+
"loss": 0.0718,
|
577 |
+
"step": 810
|
578 |
+
},
|
579 |
+
{
|
580 |
+
"epoch": 1.6734693877551021,
|
581 |
+
"grad_norm": 0.554286539554596,
|
582 |
+
"learning_rate": 9.629501659304096e-05,
|
583 |
+
"loss": 0.0687,
|
584 |
+
"step": 820
|
585 |
+
},
|
586 |
+
{
|
587 |
+
"epoch": 1.693877551020408,
|
588 |
+
"grad_norm": 0.5010990500450134,
|
589 |
+
"learning_rate": 9.61664882030862e-05,
|
590 |
+
"loss": 0.0717,
|
591 |
+
"step": 830
|
592 |
+
},
|
593 |
+
{
|
594 |
+
"epoch": 1.7142857142857144,
|
595 |
+
"grad_norm": 0.9288004040718079,
|
596 |
+
"learning_rate": 9.60358570731802e-05,
|
597 |
+
"loss": 0.068,
|
598 |
+
"step": 840
|
599 |
+
},
|
600 |
+
{
|
601 |
+
"epoch": 1.7346938775510203,
|
602 |
+
"grad_norm": 0.7913525104522705,
|
603 |
+
"learning_rate": 9.59031291531647e-05,
|
604 |
+
"loss": 0.0753,
|
605 |
+
"step": 850
|
606 |
+
},
|
607 |
+
{
|
608 |
+
"epoch": 1.7551020408163265,
|
609 |
+
"grad_norm": 0.6899034380912781,
|
610 |
+
"learning_rate": 9.57683104883836e-05,
|
611 |
+
"loss": 0.0667,
|
612 |
+
"step": 860
|
613 |
+
},
|
614 |
+
{
|
615 |
+
"epoch": 1.7755102040816326,
|
616 |
+
"grad_norm": 0.4607239365577698,
|
617 |
+
"learning_rate": 9.563140721940779e-05,
|
618 |
+
"loss": 0.068,
|
619 |
+
"step": 870
|
620 |
+
},
|
621 |
+
{
|
622 |
+
"epoch": 1.7959183673469388,
|
623 |
+
"grad_norm": 0.7755603790283203,
|
624 |
+
"learning_rate": 9.549242558175536e-05,
|
625 |
+
"loss": 0.065,
|
626 |
+
"step": 880
|
627 |
+
},
|
628 |
+
{
|
629 |
+
"epoch": 1.816326530612245,
|
630 |
+
"grad_norm": 0.7670177817344666,
|
631 |
+
"learning_rate": 9.535137190560766e-05,
|
632 |
+
"loss": 0.0641,
|
633 |
+
"step": 890
|
634 |
+
},
|
635 |
+
{
|
636 |
+
"epoch": 1.836734693877551,
|
637 |
+
"grad_norm": 0.7770594954490662,
|
638 |
+
"learning_rate": 9.520825261552092e-05,
|
639 |
+
"loss": 0.0734,
|
640 |
+
"step": 900
|
641 |
+
},
|
642 |
+
{
|
643 |
+
"epoch": 1.8571428571428572,
|
644 |
+
"grad_norm": 0.5837225317955017,
|
645 |
+
"learning_rate": 9.50630742301337e-05,
|
646 |
+
"loss": 0.0613,
|
647 |
+
"step": 910
|
648 |
+
},
|
649 |
+
{
|
650 |
+
"epoch": 1.8775510204081631,
|
651 |
+
"grad_norm": 0.7210603952407837,
|
652 |
+
"learning_rate": 9.491584336186989e-05,
|
653 |
+
"loss": 0.0688,
|
654 |
+
"step": 920
|
655 |
+
},
|
656 |
+
{
|
657 |
+
"epoch": 1.8979591836734695,
|
658 |
+
"grad_norm": 0.38964253664016724,
|
659 |
+
"learning_rate": 9.476656671663765e-05,
|
660 |
+
"loss": 0.0647,
|
661 |
+
"step": 930
|
662 |
+
},
|
663 |
+
{
|
664 |
+
"epoch": 1.9183673469387754,
|
665 |
+
"grad_norm": 0.8442269563674927,
|
666 |
+
"learning_rate": 9.46152510935239e-05,
|
667 |
+
"loss": 0.0641,
|
668 |
+
"step": 940
|
669 |
+
},
|
670 |
+
{
|
671 |
+
"epoch": 1.9387755102040818,
|
672 |
+
"grad_norm": 0.8859763741493225,
|
673 |
+
"learning_rate": 9.446190338448463e-05,
|
674 |
+
"loss": 0.0718,
|
675 |
+
"step": 950
|
676 |
+
},
|
677 |
+
{
|
678 |
+
"epoch": 1.9591836734693877,
|
679 |
+
"grad_norm": 0.670879065990448,
|
680 |
+
"learning_rate": 9.430653057403105e-05,
|
681 |
+
"loss": 0.0608,
|
682 |
+
"step": 960
|
683 |
+
},
|
684 |
+
{
|
685 |
+
"epoch": 1.9795918367346939,
|
686 |
+
"grad_norm": 0.5757398009300232,
|
687 |
+
"learning_rate": 9.414913973891144e-05,
|
688 |
+
"loss": 0.0581,
|
689 |
+
"step": 970
|
690 |
+
},
|
691 |
+
{
|
692 |
+
"epoch": 2.0,
|
693 |
+
"grad_norm": 0.6478833556175232,
|
694 |
+
"learning_rate": 9.398973804778882e-05,
|
695 |
+
"loss": 0.0636,
|
696 |
+
"step": 980
|
697 |
+
},
|
698 |
+
{
|
699 |
+
"epoch": 2.020408163265306,
|
700 |
+
"grad_norm": 0.46475401520729065,
|
701 |
+
"learning_rate": 9.382833276091447e-05,
|
702 |
+
"loss": 0.0616,
|
703 |
+
"step": 990
|
704 |
+
},
|
705 |
+
{
|
706 |
+
"epoch": 2.0408163265306123,
|
707 |
+
"grad_norm": 0.8922365307807922,
|
708 |
+
"learning_rate": 9.366493122979718e-05,
|
709 |
+
"loss": 0.0684,
|
710 |
+
"step": 1000
|
711 |
+
},
|
712 |
+
{
|
713 |
+
"epoch": 2.061224489795918,
|
714 |
+
"grad_norm": 0.5000820755958557,
|
715 |
+
"learning_rate": 9.349954089686852e-05,
|
716 |
+
"loss": 0.0563,
|
717 |
+
"step": 1010
|
718 |
+
},
|
719 |
+
{
|
720 |
+
"epoch": 2.0816326530612246,
|
721 |
+
"grad_norm": 0.5674407482147217,
|
722 |
+
"learning_rate": 9.333216929514372e-05,
|
723 |
+
"loss": 0.0558,
|
724 |
+
"step": 1020
|
725 |
+
},
|
726 |
+
{
|
727 |
+
"epoch": 2.1020408163265305,
|
728 |
+
"grad_norm": 0.7039157748222351,
|
729 |
+
"learning_rate": 9.316282404787871e-05,
|
730 |
+
"loss": 0.0653,
|
731 |
+
"step": 1030
|
732 |
+
},
|
733 |
+
{
|
734 |
+
"epoch": 2.122448979591837,
|
735 |
+
"grad_norm": 0.56775963306427,
|
736 |
+
"learning_rate": 9.299151286822278e-05,
|
737 |
+
"loss": 0.0661,
|
738 |
+
"step": 1040
|
739 |
+
},
|
740 |
+
{
|
741 |
+
"epoch": 2.142857142857143,
|
742 |
+
"grad_norm": 0.6371131539344788,
|
743 |
+
"learning_rate": 9.281824355886737e-05,
|
744 |
+
"loss": 0.0594,
|
745 |
+
"step": 1050
|
746 |
+
},
|
747 |
+
{
|
748 |
+
"epoch": 2.163265306122449,
|
749 |
+
"grad_norm": 0.6409271359443665,
|
750 |
+
"learning_rate": 9.264302401169063e-05,
|
751 |
+
"loss": 0.0658,
|
752 |
+
"step": 1060
|
753 |
+
},
|
754 |
+
{
|
755 |
+
"epoch": 2.183673469387755,
|
756 |
+
"grad_norm": 0.3617221415042877,
|
757 |
+
"learning_rate": 9.246586220739794e-05,
|
758 |
+
"loss": 0.0631,
|
759 |
+
"step": 1070
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 2.204081632653061,
|
763 |
+
"grad_norm": 0.5386372804641724,
|
764 |
+
"learning_rate": 9.228676621515853e-05,
|
765 |
+
"loss": 0.0649,
|
766 |
+
"step": 1080
|
767 |
+
},
|
768 |
+
{
|
769 |
+
"epoch": 2.2244897959183674,
|
770 |
+
"grad_norm": 0.6013124585151672,
|
771 |
+
"learning_rate": 9.210574419223777e-05,
|
772 |
+
"loss": 0.0617,
|
773 |
+
"step": 1090
|
774 |
+
},
|
775 |
+
{
|
776 |
+
"epoch": 2.2448979591836733,
|
777 |
+
"grad_norm": 0.5672731399536133,
|
778 |
+
"learning_rate": 9.192280438362582e-05,
|
779 |
+
"loss": 0.0657,
|
780 |
+
"step": 1100
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"epoch": 2.2653061224489797,
|
784 |
+
"grad_norm": 0.7263081669807434,
|
785 |
+
"learning_rate": 9.173795512166197e-05,
|
786 |
+
"loss": 0.0662,
|
787 |
+
"step": 1110
|
788 |
+
},
|
789 |
+
{
|
790 |
+
"epoch": 2.2857142857142856,
|
791 |
+
"grad_norm": 0.6938877701759338,
|
792 |
+
"learning_rate": 9.155120482565521e-05,
|
793 |
+
"loss": 0.0644,
|
794 |
+
"step": 1120
|
795 |
+
},
|
796 |
+
{
|
797 |
+
"epoch": 2.306122448979592,
|
798 |
+
"grad_norm": 0.40912479162216187,
|
799 |
+
"learning_rate": 9.136256200150066e-05,
|
800 |
+
"loss": 0.0678,
|
801 |
+
"step": 1130
|
802 |
+
},
|
803 |
+
{
|
804 |
+
"epoch": 2.326530612244898,
|
805 |
+
"grad_norm": 0.6514697074890137,
|
806 |
+
"learning_rate": 9.117203524129228e-05,
|
807 |
+
"loss": 0.0623,
|
808 |
+
"step": 1140
|
809 |
+
},
|
810 |
+
{
|
811 |
+
"epoch": 2.3469387755102042,
|
812 |
+
"grad_norm": 0.5389600992202759,
|
813 |
+
"learning_rate": 9.097963322293142e-05,
|
814 |
+
"loss": 0.0648,
|
815 |
+
"step": 1150
|
816 |
+
},
|
817 |
+
{
|
818 |
+
"epoch": 2.36734693877551,
|
819 |
+
"grad_norm": 0.5755404829978943,
|
820 |
+
"learning_rate": 9.078536470973158e-05,
|
821 |
+
"loss": 0.0604,
|
822 |
+
"step": 1160
|
823 |
+
},
|
824 |
+
{
|
825 |
+
"epoch": 2.387755102040816,
|
826 |
+
"grad_norm": 0.7156921029090881,
|
827 |
+
"learning_rate": 9.058923855001935e-05,
|
828 |
+
"loss": 0.0587,
|
829 |
+
"step": 1170
|
830 |
+
},
|
831 |
+
{
|
832 |
+
"epoch": 2.4081632653061225,
|
833 |
+
"grad_norm": 0.6002137660980225,
|
834 |
+
"learning_rate": 9.039126367673132e-05,
|
835 |
+
"loss": 0.0579,
|
836 |
+
"step": 1180
|
837 |
+
},
|
838 |
+
{
|
839 |
+
"epoch": 2.4285714285714284,
|
840 |
+
"grad_norm": 0.4457235038280487,
|
841 |
+
"learning_rate": 9.01914491070072e-05,
|
842 |
+
"loss": 0.0586,
|
843 |
+
"step": 1190
|
844 |
+
},
|
845 |
+
{
|
846 |
+
"epoch": 2.4489795918367347,
|
847 |
+
"grad_norm": 0.522216796875,
|
848 |
+
"learning_rate": 8.99898039417792e-05,
|
849 |
+
"loss": 0.0685,
|
850 |
+
"step": 1200
|
851 |
+
},
|
852 |
+
{
|
853 |
+
"epoch": 2.4693877551020407,
|
854 |
+
"grad_norm": 0.4636695683002472,
|
855 |
+
"learning_rate": 8.978633736535746e-05,
|
856 |
+
"loss": 0.0577,
|
857 |
+
"step": 1210
|
858 |
+
},
|
859 |
+
{
|
860 |
+
"epoch": 2.489795918367347,
|
861 |
+
"grad_norm": 0.40519633889198303,
|
862 |
+
"learning_rate": 8.95810586450117e-05,
|
863 |
+
"loss": 0.0532,
|
864 |
+
"step": 1220
|
865 |
+
},
|
866 |
+
{
|
867 |
+
"epoch": 2.510204081632653,
|
868 |
+
"grad_norm": 0.7298558950424194,
|
869 |
+
"learning_rate": 8.937397713054916e-05,
|
870 |
+
"loss": 0.0646,
|
871 |
+
"step": 1230
|
872 |
+
},
|
873 |
+
{
|
874 |
+
"epoch": 2.5306122448979593,
|
875 |
+
"grad_norm": 0.4737094044685364,
|
876 |
+
"learning_rate": 8.916510225388878e-05,
|
877 |
+
"loss": 0.0627,
|
878 |
+
"step": 1240
|
879 |
+
},
|
880 |
+
{
|
881 |
+
"epoch": 2.5510204081632653,
|
882 |
+
"grad_norm": 0.3282434940338135,
|
883 |
+
"learning_rate": 8.895444352863154e-05,
|
884 |
+
"loss": 0.0617,
|
885 |
+
"step": 1250
|
886 |
+
},
|
887 |
+
{
|
888 |
+
"epoch": 2.571428571428571,
|
889 |
+
"grad_norm": 0.6259459257125854,
|
890 |
+
"learning_rate": 8.874201054962721e-05,
|
891 |
+
"loss": 0.0628,
|
892 |
+
"step": 1260
|
893 |
+
},
|
894 |
+
{
|
895 |
+
"epoch": 2.5918367346938775,
|
896 |
+
"grad_norm": 0.5925482511520386,
|
897 |
+
"learning_rate": 8.852781299253724e-05,
|
898 |
+
"loss": 0.0618,
|
899 |
+
"step": 1270
|
900 |
+
},
|
901 |
+
{
|
902 |
+
"epoch": 2.612244897959184,
|
903 |
+
"grad_norm": 0.4080643951892853,
|
904 |
+
"learning_rate": 8.83118606133942e-05,
|
905 |
+
"loss": 0.0587,
|
906 |
+
"step": 1280
|
907 |
+
},
|
908 |
+
{
|
909 |
+
"epoch": 2.63265306122449,
|
910 |
+
"grad_norm": 0.5757149457931519,
|
911 |
+
"learning_rate": 8.809416324815729e-05,
|
912 |
+
"loss": 0.0647,
|
913 |
+
"step": 1290
|
914 |
+
},
|
915 |
+
{
|
916 |
+
"epoch": 2.6530612244897958,
|
917 |
+
"grad_norm": 0.6882585287094116,
|
918 |
+
"learning_rate": 8.78747308122644e-05,
|
919 |
+
"loss": 0.0611,
|
920 |
+
"step": 1300
|
921 |
+
},
|
922 |
+
{
|
923 |
+
"epoch": 2.673469387755102,
|
924 |
+
"grad_norm": 0.4955786466598511,
|
925 |
+
"learning_rate": 8.765357330018056e-05,
|
926 |
+
"loss": 0.0595,
|
927 |
+
"step": 1310
|
928 |
+
},
|
929 |
+
{
|
930 |
+
"epoch": 2.693877551020408,
|
931 |
+
"grad_norm": 0.425704687833786,
|
932 |
+
"learning_rate": 8.743070078494255e-05,
|
933 |
+
"loss": 0.069,
|
934 |
+
"step": 1320
|
935 |
+
},
|
936 |
+
{
|
937 |
+
"epoch": 2.7142857142857144,
|
938 |
+
"grad_norm": 0.6342069506645203,
|
939 |
+
"learning_rate": 8.720612341770027e-05,
|
940 |
+
"loss": 0.0605,
|
941 |
+
"step": 1330
|
942 |
+
},
|
943 |
+
{
|
944 |
+
"epoch": 2.7346938775510203,
|
945 |
+
"grad_norm": 0.5021179914474487,
|
946 |
+
"learning_rate": 8.697985142725435e-05,
|
947 |
+
"loss": 0.0632,
|
948 |
+
"step": 1340
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"epoch": 2.7551020408163263,
|
952 |
+
"grad_norm": 0.4925266206264496,
|
953 |
+
"learning_rate": 8.67518951195902e-05,
|
954 |
+
"loss": 0.0598,
|
955 |
+
"step": 1350
|
956 |
+
},
|
957 |
+
{
|
958 |
+
"epoch": 2.7755102040816326,
|
959 |
+
"grad_norm": 0.5765326023101807,
|
960 |
+
"learning_rate": 8.652226487740863e-05,
|
961 |
+
"loss": 0.0577,
|
962 |
+
"step": 1360
|
963 |
+
},
|
964 |
+
{
|
965 |
+
"epoch": 2.795918367346939,
|
966 |
+
"grad_norm": 0.8558735847473145,
|
967 |
+
"learning_rate": 8.629097115965298e-05,
|
968 |
+
"loss": 0.0614,
|
969 |
+
"step": 1370
|
970 |
+
},
|
971 |
+
{
|
972 |
+
"epoch": 2.816326530612245,
|
973 |
+
"grad_norm": 0.6366243958473206,
|
974 |
+
"learning_rate": 8.605802450103275e-05,
|
975 |
+
"loss": 0.0692,
|
976 |
+
"step": 1380
|
977 |
+
},
|
978 |
+
{
|
979 |
+
"epoch": 2.836734693877551,
|
980 |
+
"grad_norm": 0.97010737657547,
|
981 |
+
"learning_rate": 8.58234355115437e-05,
|
982 |
+
"loss": 0.0567,
|
983 |
+
"step": 1390
|
984 |
+
},
|
985 |
+
{
|
986 |
+
"epoch": 2.857142857142857,
|
987 |
+
"grad_norm": 0.48445919156074524,
|
988 |
+
"learning_rate": 8.558721487598471e-05,
|
989 |
+
"loss": 0.054,
|
990 |
+
"step": 1400
|
991 |
+
},
|
992 |
+
{
|
993 |
+
"epoch": 2.877551020408163,
|
994 |
+
"grad_norm": 0.5952962040901184,
|
995 |
+
"learning_rate": 8.534937335347102e-05,
|
996 |
+
"loss": 0.0529,
|
997 |
+
"step": 1410
|
998 |
+
},
|
999 |
+
{
|
1000 |
+
"epoch": 2.8979591836734695,
|
1001 |
+
"grad_norm": 0.5294446349143982,
|
1002 |
+
"learning_rate": 8.510992177694429e-05,
|
1003 |
+
"loss": 0.055,
|
1004 |
+
"step": 1420
|
1005 |
+
},
|
1006 |
+
{
|
1007 |
+
"epoch": 2.9183673469387754,
|
1008 |
+
"grad_norm": 0.47300493717193604,
|
1009 |
+
"learning_rate": 8.486887105267902e-05,
|
1010 |
+
"loss": 0.0598,
|
1011 |
+
"step": 1430
|
1012 |
+
},
|
1013 |
+
{
|
1014 |
+
"epoch": 2.938775510204082,
|
1015 |
+
"grad_norm": 0.37820449471473694,
|
1016 |
+
"learning_rate": 8.462623215978605e-05,
|
1017 |
+
"loss": 0.0618,
|
1018 |
+
"step": 1440
|
1019 |
+
},
|
1020 |
+
{
|
1021 |
+
"epoch": 2.9591836734693877,
|
1022 |
+
"grad_norm": 0.46813079714775085,
|
1023 |
+
"learning_rate": 8.438201614971226e-05,
|
1024 |
+
"loss": 0.059,
|
1025 |
+
"step": 1450
|
1026 |
+
},
|
1027 |
+
{
|
1028 |
+
"epoch": 2.979591836734694,
|
1029 |
+
"grad_norm": 0.5875042080879211,
|
1030 |
+
"learning_rate": 8.41362341457374e-05,
|
1031 |
+
"loss": 0.0564,
|
1032 |
+
"step": 1460
|
1033 |
+
},
|
1034 |
+
{
|
1035 |
+
"epoch": 3.0,
|
1036 |
+
"grad_norm": 0.7085863947868347,
|
1037 |
+
"learning_rate": 8.388889734246725e-05,
|
1038 |
+
"loss": 0.058,
|
1039 |
+
"step": 1470
|
1040 |
+
},
|
1041 |
+
{
|
1042 |
+
"epoch": 3.020408163265306,
|
1043 |
+
"grad_norm": 0.5409305095672607,
|
1044 |
+
"learning_rate": 8.364001700532401e-05,
|
1045 |
+
"loss": 0.0666,
|
1046 |
+
"step": 1480
|
1047 |
+
},
|
1048 |
+
{
|
1049 |
+
"epoch": 3.0408163265306123,
|
1050 |
+
"grad_norm": 0.7195011377334595,
|
1051 |
+
"learning_rate": 8.338960447003292e-05,
|
1052 |
+
"loss": 0.0578,
|
1053 |
+
"step": 1490
|
1054 |
+
},
|
1055 |
+
{
|
1056 |
+
"epoch": 3.061224489795918,
|
1057 |
+
"grad_norm": 0.4446586072444916,
|
1058 |
+
"learning_rate": 8.313767114210615e-05,
|
1059 |
+
"loss": 0.0634,
|
1060 |
+
"step": 1500
|
1061 |
+
},
|
1062 |
+
{
|
1063 |
+
"epoch": 3.0816326530612246,
|
1064 |
+
"grad_norm": 0.4142863154411316,
|
1065 |
+
"learning_rate": 8.288422849632325e-05,
|
1066 |
+
"loss": 0.0652,
|
1067 |
+
"step": 1510
|
1068 |
+
},
|
1069 |
+
{
|
1070 |
+
"epoch": 3.1020408163265305,
|
1071 |
+
"grad_norm": 0.47294557094573975,
|
1072 |
+
"learning_rate": 8.262928807620843e-05,
|
1073 |
+
"loss": 0.0552,
|
1074 |
+
"step": 1520
|
1075 |
+
},
|
1076 |
+
{
|
1077 |
+
"epoch": 3.122448979591837,
|
1078 |
+
"grad_norm": 0.4253334403038025,
|
1079 |
+
"learning_rate": 8.237286149350495e-05,
|
1080 |
+
"loss": 0.0596,
|
1081 |
+
"step": 1530
|
1082 |
+
},
|
1083 |
+
{
|
1084 |
+
"epoch": 3.142857142857143,
|
1085 |
+
"grad_norm": 0.4108365774154663,
|
1086 |
+
"learning_rate": 8.211496042764612e-05,
|
1087 |
+
"loss": 0.059,
|
1088 |
+
"step": 1540
|
1089 |
+
},
|
1090 |
+
{
|
1091 |
+
"epoch": 3.163265306122449,
|
1092 |
+
"grad_norm": 0.4678657650947571,
|
1093 |
+
"learning_rate": 8.185559662522337e-05,
|
1094 |
+
"loss": 0.0617,
|
1095 |
+
"step": 1550
|
1096 |
+
},
|
1097 |
+
{
|
1098 |
+
"epoch": 3.183673469387755,
|
1099 |
+
"grad_norm": 0.5301884412765503,
|
1100 |
+
"learning_rate": 8.159478189945121e-05,
|
1101 |
+
"loss": 0.0513,
|
1102 |
+
"step": 1560
|
1103 |
+
},
|
1104 |
+
{
|
1105 |
+
"epoch": 3.204081632653061,
|
1106 |
+
"grad_norm": 0.8106349110603333,
|
1107 |
+
"learning_rate": 8.133252812962922e-05,
|
1108 |
+
"loss": 0.0567,
|
1109 |
+
"step": 1570
|
1110 |
+
},
|
1111 |
+
{
|
1112 |
+
"epoch": 3.2244897959183674,
|
1113 |
+
"grad_norm": 0.4195539355278015,
|
1114 |
+
"learning_rate": 8.106884726060098e-05,
|
1115 |
+
"loss": 0.053,
|
1116 |
+
"step": 1580
|
1117 |
+
},
|
1118 |
+
{
|
1119 |
+
"epoch": 3.2448979591836733,
|
1120 |
+
"grad_norm": 0.5291970372200012,
|
1121 |
+
"learning_rate": 8.080375130220994e-05,
|
1122 |
+
"loss": 0.0549,
|
1123 |
+
"step": 1590
|
1124 |
+
},
|
1125 |
+
{
|
1126 |
+
"epoch": 3.2653061224489797,
|
1127 |
+
"grad_norm": 0.45250171422958374,
|
1128 |
+
"learning_rate": 8.053725232875253e-05,
|
1129 |
+
"loss": 0.0563,
|
1130 |
+
"step": 1600
|
1131 |
+
},
|
1132 |
+
{
|
1133 |
+
"epoch": 3.2857142857142856,
|
1134 |
+
"grad_norm": 0.5225273966789246,
|
1135 |
+
"learning_rate": 8.026936247842813e-05,
|
1136 |
+
"loss": 0.0547,
|
1137 |
+
"step": 1610
|
1138 |
+
},
|
1139 |
+
{
|
1140 |
+
"epoch": 3.306122448979592,
|
1141 |
+
"grad_norm": 0.4152168929576874,
|
1142 |
+
"learning_rate": 8.000009395278624e-05,
|
1143 |
+
"loss": 0.0491,
|
1144 |
+
"step": 1620
|
1145 |
+
},
|
1146 |
+
{
|
1147 |
+
"epoch": 3.326530612244898,
|
1148 |
+
"grad_norm": 0.4513435959815979,
|
1149 |
+
"learning_rate": 7.972945901617072e-05,
|
1150 |
+
"loss": 0.0522,
|
1151 |
+
"step": 1630
|
1152 |
+
},
|
1153 |
+
{
|
1154 |
+
"epoch": 3.3469387755102042,
|
1155 |
+
"grad_norm": 0.38231614232063293,
|
1156 |
+
"learning_rate": 7.945746999516119e-05,
|
1157 |
+
"loss": 0.0496,
|
1158 |
+
"step": 1640
|
1159 |
+
},
|
1160 |
+
{
|
1161 |
+
"epoch": 3.36734693877551,
|
1162 |
+
"grad_norm": 0.47759726643562317,
|
1163 |
+
"learning_rate": 7.918413927801164e-05,
|
1164 |
+
"loss": 0.0555,
|
1165 |
+
"step": 1650
|
1166 |
+
},
|
1167 |
+
{
|
1168 |
+
"epoch": 3.387755102040816,
|
1169 |
+
"grad_norm": 0.47755759954452515,
|
1170 |
+
"learning_rate": 7.890947931408613e-05,
|
1171 |
+
"loss": 0.0468,
|
1172 |
+
"step": 1660
|
1173 |
+
},
|
1174 |
+
{
|
1175 |
+
"epoch": 3.4081632653061225,
|
1176 |
+
"grad_norm": 0.34949928522109985,
|
1177 |
+
"learning_rate": 7.863350261329177e-05,
|
1178 |
+
"loss": 0.0459,
|
1179 |
+
"step": 1670
|
1180 |
+
},
|
1181 |
+
{
|
1182 |
+
"epoch": 3.4285714285714284,
|
1183 |
+
"grad_norm": 0.39500531554222107,
|
1184 |
+
"learning_rate": 7.835622174550891e-05,
|
1185 |
+
"loss": 0.0545,
|
1186 |
+
"step": 1680
|
1187 |
+
},
|
1188 |
+
{
|
1189 |
+
"epoch": 3.4489795918367347,
|
1190 |
+
"grad_norm": 0.3671438992023468,
|
1191 |
+
"learning_rate": 7.807764934001874e-05,
|
1192 |
+
"loss": 0.0542,
|
1193 |
+
"step": 1690
|
1194 |
+
},
|
1195 |
+
{
|
1196 |
+
"epoch": 3.4693877551020407,
|
1197 |
+
"grad_norm": 0.4870457351207733,
|
1198 |
+
"learning_rate": 7.779779808492788e-05,
|
1199 |
+
"loss": 0.0525,
|
1200 |
+
"step": 1700
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"epoch": 3.489795918367347,
|
1204 |
+
"grad_norm": 0.6565643548965454,
|
1205 |
+
"learning_rate": 7.751668072659061e-05,
|
1206 |
+
"loss": 0.0559,
|
1207 |
+
"step": 1710
|
1208 |
+
},
|
1209 |
+
{
|
1210 |
+
"epoch": 3.510204081632653,
|
1211 |
+
"grad_norm": 0.4305172860622406,
|
1212 |
+
"learning_rate": 7.723431006902829e-05,
|
1213 |
+
"loss": 0.0516,
|
1214 |
+
"step": 1720
|
1215 |
+
},
|
1216 |
+
{
|
1217 |
+
"epoch": 3.5306122448979593,
|
1218 |
+
"grad_norm": 0.44627729058265686,
|
1219 |
+
"learning_rate": 7.695069897334612e-05,
|
1220 |
+
"loss": 0.0495,
|
1221 |
+
"step": 1730
|
1222 |
+
},
|
1223 |
+
{
|
1224 |
+
"epoch": 3.5510204081632653,
|
1225 |
+
"grad_norm": 0.39987713098526,
|
1226 |
+
"learning_rate": 7.666586035714745e-05,
|
1227 |
+
"loss": 0.0521,
|
1228 |
+
"step": 1740
|
1229 |
+
},
|
1230 |
+
{
|
1231 |
+
"epoch": 3.571428571428571,
|
1232 |
+
"grad_norm": 0.39626452326774597,
|
1233 |
+
"learning_rate": 7.637980719394533e-05,
|
1234 |
+
"loss": 0.0605,
|
1235 |
+
"step": 1750
|
1236 |
+
},
|
1237 |
+
{
|
1238 |
+
"epoch": 3.5918367346938775,
|
1239 |
+
"grad_norm": 0.7232034206390381,
|
1240 |
+
"learning_rate": 7.609255251257168e-05,
|
1241 |
+
"loss": 0.0626,
|
1242 |
+
"step": 1760
|
1243 |
+
},
|
1244 |
+
{
|
1245 |
+
"epoch": 3.612244897959184,
|
1246 |
+
"grad_norm": 0.4168018102645874,
|
1247 |
+
"learning_rate": 7.580410939658383e-05,
|
1248 |
+
"loss": 0.0525,
|
1249 |
+
"step": 1770
|
1250 |
+
},
|
1251 |
+
{
|
1252 |
+
"epoch": 3.63265306122449,
|
1253 |
+
"grad_norm": 0.45826172828674316,
|
1254 |
+
"learning_rate": 7.551449098366857e-05,
|
1255 |
+
"loss": 0.0587,
|
1256 |
+
"step": 1780
|
1257 |
+
},
|
1258 |
+
{
|
1259 |
+
"epoch": 3.6530612244897958,
|
1260 |
+
"grad_norm": 0.5224451422691345,
|
1261 |
+
"learning_rate": 7.522371046504385e-05,
|
1262 |
+
"loss": 0.0541,
|
1263 |
+
"step": 1790
|
1264 |
+
},
|
1265 |
+
{
|
1266 |
+
"epoch": 3.673469387755102,
|
1267 |
+
"grad_norm": 0.577448308467865,
|
1268 |
+
"learning_rate": 7.493178108485792e-05,
|
1269 |
+
"loss": 0.0591,
|
1270 |
+
"step": 1800
|
1271 |
+
},
|
1272 |
+
{
|
1273 |
+
"epoch": 3.693877551020408,
|
1274 |
+
"grad_norm": 0.5018607974052429,
|
1275 |
+
"learning_rate": 7.463871613958608e-05,
|
1276 |
+
"loss": 0.0466,
|
1277 |
+
"step": 1810
|
1278 |
+
},
|
1279 |
+
{
|
1280 |
+
"epoch": 3.7142857142857144,
|
1281 |
+
"grad_norm": 0.38134491443634033,
|
1282 |
+
"learning_rate": 7.434452897742513e-05,
|
1283 |
+
"loss": 0.0527,
|
1284 |
+
"step": 1820
|
1285 |
+
},
|
1286 |
+
{
|
1287 |
+
"epoch": 3.7346938775510203,
|
1288 |
+
"grad_norm": 0.4851766526699066,
|
1289 |
+
"learning_rate": 7.40492329976853e-05,
|
1290 |
+
"loss": 0.0524,
|
1291 |
+
"step": 1830
|
1292 |
+
},
|
1293 |
+
{
|
1294 |
+
"epoch": 3.7551020408163263,
|
1295 |
+
"grad_norm": 0.4203612804412842,
|
1296 |
+
"learning_rate": 7.375284165018003e-05,
|
1297 |
+
"loss": 0.0548,
|
1298 |
+
"step": 1840
|
1299 |
+
},
|
1300 |
+
{
|
1301 |
+
"epoch": 3.7755102040816326,
|
1302 |
+
"grad_norm": 0.41804400086402893,
|
1303 |
+
"learning_rate": 7.345536843461342e-05,
|
1304 |
+
"loss": 0.0526,
|
1305 |
+
"step": 1850
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"epoch": 3.795918367346939,
|
1309 |
+
"grad_norm": 0.45135030150413513,
|
1310 |
+
"learning_rate": 7.31568268999652e-05,
|
1311 |
+
"loss": 0.0602,
|
1312 |
+
"step": 1860
|
1313 |
+
},
|
1314 |
+
{
|
1315 |
+
"epoch": 3.816326530612245,
|
1316 |
+
"grad_norm": 0.4034428894519806,
|
1317 |
+
"learning_rate": 7.285723064387373e-05,
|
1318 |
+
"loss": 0.0529,
|
1319 |
+
"step": 1870
|
1320 |
+
},
|
1321 |
+
{
|
1322 |
+
"epoch": 3.836734693877551,
|
1323 |
+
"grad_norm": 0.48290595412254333,
|
1324 |
+
"learning_rate": 7.255659331201673e-05,
|
1325 |
+
"loss": 0.0547,
|
1326 |
+
"step": 1880
|
1327 |
+
},
|
1328 |
+
{
|
1329 |
+
"epoch": 3.857142857142857,
|
1330 |
+
"grad_norm": 0.7128674387931824,
|
1331 |
+
"learning_rate": 7.22549285974896e-05,
|
1332 |
+
"loss": 0.0522,
|
1333 |
+
"step": 1890
|
1334 |
+
},
|
1335 |
+
{
|
1336 |
+
"epoch": 3.877551020408163,
|
1337 |
+
"grad_norm": 0.5425318479537964,
|
1338 |
+
"learning_rate": 7.195225024018187e-05,
|
1339 |
+
"loss": 0.0582,
|
1340 |
+
"step": 1900
|
1341 |
+
},
|
1342 |
+
{
|
1343 |
+
"epoch": 3.8979591836734695,
|
1344 |
+
"grad_norm": 0.4822893440723419,
|
1345 |
+
"learning_rate": 7.164857202615129e-05,
|
1346 |
+
"loss": 0.0508,
|
1347 |
+
"step": 1910
|
1348 |
+
},
|
1349 |
+
{
|
1350 |
+
"epoch": 3.9183673469387754,
|
1351 |
+
"grad_norm": 0.5935354828834534,
|
1352 |
+
"learning_rate": 7.134390778699604e-05,
|
1353 |
+
"loss": 0.0482,
|
1354 |
+
"step": 1920
|
1355 |
+
},
|
1356 |
+
{
|
1357 |
+
"epoch": 3.938775510204082,
|
1358 |
+
"grad_norm": 0.372458279132843,
|
1359 |
+
"learning_rate": 7.103827139922465e-05,
|
1360 |
+
"loss": 0.0503,
|
1361 |
+
"step": 1930
|
1362 |
+
},
|
1363 |
+
{
|
1364 |
+
"epoch": 3.9591836734693877,
|
1365 |
+
"grad_norm": 0.46793851256370544,
|
1366 |
+
"learning_rate": 7.073167678362401e-05,
|
1367 |
+
"loss": 0.0567,
|
1368 |
+
"step": 1940
|
1369 |
+
},
|
1370 |
+
{
|
1371 |
+
"epoch": 3.979591836734694,
|
1372 |
+
"grad_norm": 0.5541433095932007,
|
1373 |
+
"learning_rate": 7.042413790462528e-05,
|
1374 |
+
"loss": 0.0564,
|
1375 |
+
"step": 1950
|
1376 |
+
},
|
1377 |
+
{
|
1378 |
+
"epoch": 4.0,
|
1379 |
+
"grad_norm": 0.5569531917572021,
|
1380 |
+
"learning_rate": 7.011566876966786e-05,
|
1381 |
+
"loss": 0.0536,
|
1382 |
+
"step": 1960
|
1383 |
+
},
|
1384 |
+
{
|
1385 |
+
"epoch": 4.020408163265306,
|
1386 |
+
"grad_norm": 0.48454782366752625,
|
1387 |
+
"learning_rate": 6.980628342856152e-05,
|
1388 |
+
"loss": 0.0545,
|
1389 |
+
"step": 1970
|
1390 |
+
},
|
1391 |
+
{
|
1392 |
+
"epoch": 4.040816326530612,
|
1393 |
+
"grad_norm": 0.5605842471122742,
|
1394 |
+
"learning_rate": 6.949599597284625e-05,
|
1395 |
+
"loss": 0.0533,
|
1396 |
+
"step": 1980
|
1397 |
+
},
|
1398 |
+
{
|
1399 |
+
"epoch": 4.061224489795919,
|
1400 |
+
"grad_norm": 0.5089229941368103,
|
1401 |
+
"learning_rate": 6.91848205351506e-05,
|
1402 |
+
"loss": 0.0497,
|
1403 |
+
"step": 1990
|
1404 |
+
},
|
1405 |
+
{
|
1406 |
+
"epoch": 4.081632653061225,
|
1407 |
+
"grad_norm": 0.42554935812950134,
|
1408 |
+
"learning_rate": 6.8872771288548e-05,
|
1409 |
+
"loss": 0.0495,
|
1410 |
+
"step": 2000
|
1411 |
+
},
|
1412 |
+
{
|
1413 |
+
"epoch": 4.1020408163265305,
|
1414 |
+
"grad_norm": 0.6508166790008545,
|
1415 |
+
"learning_rate": 6.855986244591104e-05,
|
1416 |
+
"loss": 0.0499,
|
1417 |
+
"step": 2010
|
1418 |
+
},
|
1419 |
+
{
|
1420 |
+
"epoch": 4.122448979591836,
|
1421 |
+
"grad_norm": 0.46225705742836,
|
1422 |
+
"learning_rate": 6.824610825926433e-05,
|
1423 |
+
"loss": 0.0484,
|
1424 |
+
"step": 2020
|
1425 |
+
},
|
1426 |
+
{
|
1427 |
+
"epoch": 4.142857142857143,
|
1428 |
+
"grad_norm": 0.38654595613479614,
|
1429 |
+
"learning_rate": 6.793152301913523e-05,
|
1430 |
+
"loss": 0.0525,
|
1431 |
+
"step": 2030
|
1432 |
+
},
|
1433 |
+
{
|
1434 |
+
"epoch": 4.163265306122449,
|
1435 |
+
"grad_norm": 0.43045929074287415,
|
1436 |
+
"learning_rate": 6.7616121053903e-05,
|
1437 |
+
"loss": 0.0601,
|
1438 |
+
"step": 2040
|
1439 |
+
},
|
1440 |
+
{
|
1441 |
+
"epoch": 4.183673469387755,
|
1442 |
+
"grad_norm": 0.29230445623397827,
|
1443 |
+
"learning_rate": 6.72999167291462e-05,
|
1444 |
+
"loss": 0.0442,
|
1445 |
+
"step": 2050
|
1446 |
+
},
|
1447 |
+
{
|
1448 |
+
"epoch": 4.204081632653061,
|
1449 |
+
"grad_norm": 0.4340389370918274,
|
1450 |
+
"learning_rate": 6.698292444698839e-05,
|
1451 |
+
"loss": 0.0455,
|
1452 |
+
"step": 2060
|
1453 |
+
},
|
1454 |
+
{
|
1455 |
+
"epoch": 4.224489795918367,
|
1456 |
+
"grad_norm": 0.6728600859642029,
|
1457 |
+
"learning_rate": 6.666515864544209e-05,
|
1458 |
+
"loss": 0.0419,
|
1459 |
+
"step": 2070
|
1460 |
+
},
|
1461 |
+
{
|
1462 |
+
"epoch": 4.244897959183674,
|
1463 |
+
"grad_norm": 0.4668474495410919,
|
1464 |
+
"learning_rate": 6.634663379775125e-05,
|
1465 |
+
"loss": 0.0466,
|
1466 |
+
"step": 2080
|
1467 |
+
},
|
1468 |
+
{
|
1469 |
+
"epoch": 4.26530612244898,
|
1470 |
+
"grad_norm": 0.35625386238098145,
|
1471 |
+
"learning_rate": 6.602736441173204e-05,
|
1472 |
+
"loss": 0.0408,
|
1473 |
+
"step": 2090
|
1474 |
+
},
|
1475 |
+
{
|
1476 |
+
"epoch": 4.285714285714286,
|
1477 |
+
"grad_norm": 0.29537519812583923,
|
1478 |
+
"learning_rate": 6.570736502911197e-05,
|
1479 |
+
"loss": 0.0401,
|
1480 |
+
"step": 2100
|
1481 |
+
},
|
1482 |
+
{
|
1483 |
+
"epoch": 4.3061224489795915,
|
1484 |
+
"grad_norm": 0.4634284973144531,
|
1485 |
+
"learning_rate": 6.538665022486767e-05,
|
1486 |
+
"loss": 0.0476,
|
1487 |
+
"step": 2110
|
1488 |
+
},
|
1489 |
+
{
|
1490 |
+
"epoch": 4.326530612244898,
|
1491 |
+
"grad_norm": 0.3917020559310913,
|
1492 |
+
"learning_rate": 6.506523460656103e-05,
|
1493 |
+
"loss": 0.0441,
|
1494 |
+
"step": 2120
|
1495 |
+
},
|
1496 |
+
{
|
1497 |
+
"epoch": 4.346938775510204,
|
1498 |
+
"grad_norm": 0.48713311553001404,
|
1499 |
+
"learning_rate": 6.474313281367373e-05,
|
1500 |
+
"loss": 0.0481,
|
1501 |
+
"step": 2130
|
1502 |
+
},
|
1503 |
+
{
|
1504 |
+
"epoch": 4.36734693877551,
|
1505 |
+
"grad_norm": 0.4484109580516815,
|
1506 |
+
"learning_rate": 6.442035951694068e-05,
|
1507 |
+
"loss": 0.044,
|
1508 |
+
"step": 2140
|
1509 |
+
},
|
1510 |
+
{
|
1511 |
+
"epoch": 4.387755102040816,
|
1512 |
+
"grad_norm": 0.5104283690452576,
|
1513 |
+
"learning_rate": 6.409692941768166e-05,
|
1514 |
+
"loss": 0.0562,
|
1515 |
+
"step": 2150
|
1516 |
+
},
|
1517 |
+
{
|
1518 |
+
"epoch": 4.408163265306122,
|
1519 |
+
"grad_norm": 0.5283732414245605,
|
1520 |
+
"learning_rate": 6.377285724713176e-05,
|
1521 |
+
"loss": 0.045,
|
1522 |
+
"step": 2160
|
1523 |
+
},
|
1524 |
+
{
|
1525 |
+
"epoch": 4.428571428571429,
|
1526 |
+
"grad_norm": 0.4185742735862732,
|
1527 |
+
"learning_rate": 6.344815776577041e-05,
|
1528 |
+
"loss": 0.044,
|
1529 |
+
"step": 2170
|
1530 |
+
},
|
1531 |
+
{
|
1532 |
+
"epoch": 4.448979591836735,
|
1533 |
+
"grad_norm": 0.5266509056091309,
|
1534 |
+
"learning_rate": 6.312284576264913e-05,
|
1535 |
+
"loss": 0.0475,
|
1536 |
+
"step": 2180
|
1537 |
+
},
|
1538 |
+
{
|
1539 |
+
"epoch": 4.469387755102041,
|
1540 |
+
"grad_norm": 0.34054675698280334,
|
1541 |
+
"learning_rate": 6.279693605471787e-05,
|
1542 |
+
"loss": 0.0508,
|
1543 |
+
"step": 2190
|
1544 |
+
},
|
1545 |
+
{
|
1546 |
+
"epoch": 4.489795918367347,
|
1547 |
+
"grad_norm": 0.42366960644721985,
|
1548 |
+
"learning_rate": 6.24704434861502e-05,
|
1549 |
+
"loss": 0.0483,
|
1550 |
+
"step": 2200
|
1551 |
+
},
|
1552 |
+
{
|
1553 |
+
"epoch": 4.510204081632653,
|
1554 |
+
"grad_norm": 0.46188119053840637,
|
1555 |
+
"learning_rate": 6.214338292766715e-05,
|
1556 |
+
"loss": 0.0413,
|
1557 |
+
"step": 2210
|
1558 |
+
},
|
1559 |
+
{
|
1560 |
+
"epoch": 4.530612244897959,
|
1561 |
+
"grad_norm": 0.5456094741821289,
|
1562 |
+
"learning_rate": 6.181576927585992e-05,
|
1563 |
+
"loss": 0.0362,
|
1564 |
+
"step": 2220
|
1565 |
+
},
|
1566 |
+
{
|
1567 |
+
"epoch": 4.551020408163265,
|
1568 |
+
"grad_norm": 0.3864874243736267,
|
1569 |
+
"learning_rate": 6.148761745251146e-05,
|
1570 |
+
"loss": 0.0413,
|
1571 |
+
"step": 2230
|
1572 |
+
},
|
1573 |
+
{
|
1574 |
+
"epoch": 4.571428571428571,
|
1575 |
+
"grad_norm": 0.3080037534236908,
|
1576 |
+
"learning_rate": 6.115894240391666e-05,
|
1577 |
+
"loss": 0.0442,
|
1578 |
+
"step": 2240
|
1579 |
+
},
|
1580 |
+
{
|
1581 |
+
"epoch": 4.591836734693878,
|
1582 |
+
"grad_norm": 0.3093554675579071,
|
1583 |
+
"learning_rate": 6.082975910020179e-05,
|
1584 |
+
"loss": 0.0455,
|
1585 |
+
"step": 2250
|
1586 |
+
},
|
1587 |
+
{
|
1588 |
+
"epoch": 4.612244897959184,
|
1589 |
+
"grad_norm": 0.3666916489601135,
|
1590 |
+
"learning_rate": 6.0500082534642464e-05,
|
1591 |
+
"loss": 0.0465,
|
1592 |
+
"step": 2260
|
1593 |
+
},
|
1594 |
+
{
|
1595 |
+
"epoch": 4.63265306122449,
|
1596 |
+
"grad_norm": 0.3994399607181549,
|
1597 |
+
"learning_rate": 6.0169927722980935e-05,
|
1598 |
+
"loss": 0.045,
|
1599 |
+
"step": 2270
|
1600 |
+
},
|
1601 |
+
{
|
1602 |
+
"epoch": 4.653061224489796,
|
1603 |
+
"grad_norm": 0.4581643044948578,
|
1604 |
+
"learning_rate": 5.9839309702741995e-05,
|
1605 |
+
"loss": 0.0477,
|
1606 |
+
"step": 2280
|
1607 |
+
},
|
1608 |
+
{
|
1609 |
+
"epoch": 4.673469387755102,
|
1610 |
+
"grad_norm": 0.47960394620895386,
|
1611 |
+
"learning_rate": 5.950824353254818e-05,
|
1612 |
+
"loss": 0.0451,
|
1613 |
+
"step": 2290
|
1614 |
+
},
|
1615 |
+
{
|
1616 |
+
"epoch": 4.6938775510204085,
|
1617 |
+
"grad_norm": 0.4794948399066925,
|
1618 |
+
"learning_rate": 5.9176744291433895e-05,
|
1619 |
+
"loss": 0.0426,
|
1620 |
+
"step": 2300
|
1621 |
+
},
|
1622 |
+
{
|
1623 |
+
"epoch": 4.714285714285714,
|
1624 |
+
"grad_norm": 0.39106351137161255,
|
1625 |
+
"learning_rate": 5.8844827078158525e-05,
|
1626 |
+
"loss": 0.0473,
|
1627 |
+
"step": 2310
|
1628 |
+
},
|
1629 |
+
{
|
1630 |
+
"epoch": 4.73469387755102,
|
1631 |
+
"grad_norm": 0.48323217034339905,
|
1632 |
+
"learning_rate": 5.851250701051881e-05,
|
1633 |
+
"loss": 0.0439,
|
1634 |
+
"step": 2320
|
1635 |
+
},
|
1636 |
+
{
|
1637 |
+
"epoch": 4.755102040816326,
|
1638 |
+
"grad_norm": 0.397773802280426,
|
1639 |
+
"learning_rate": 5.817979922466028e-05,
|
1640 |
+
"loss": 0.0481,
|
1641 |
+
"step": 2330
|
1642 |
+
},
|
1643 |
+
{
|
1644 |
+
"epoch": 4.775510204081632,
|
1645 |
+
"grad_norm": 0.9115492105484009,
|
1646 |
+
"learning_rate": 5.784671887438775e-05,
|
1647 |
+
"loss": 0.047,
|
1648 |
+
"step": 2340
|
1649 |
+
},
|
1650 |
+
{
|
1651 |
+
"epoch": 4.795918367346939,
|
1652 |
+
"grad_norm": 0.4305925965309143,
|
1653 |
+
"learning_rate": 5.751328113047527e-05,
|
1654 |
+
"loss": 0.0492,
|
1655 |
+
"step": 2350
|
1656 |
+
},
|
1657 |
+
{
|
1658 |
+
"epoch": 4.816326530612245,
|
1659 |
+
"grad_norm": 0.43962588906288147,
|
1660 |
+
"learning_rate": 5.717950117997501e-05,
|
1661 |
+
"loss": 0.0457,
|
1662 |
+
"step": 2360
|
1663 |
+
},
|
1664 |
+
{
|
1665 |
+
"epoch": 4.836734693877551,
|
1666 |
+
"grad_norm": 0.47304484248161316,
|
1667 |
+
"learning_rate": 5.6845394225525605e-05,
|
1668 |
+
"loss": 0.0435,
|
1669 |
+
"step": 2370
|
1670 |
+
},
|
1671 |
+
{
|
1672 |
+
"epoch": 4.857142857142857,
|
1673 |
+
"grad_norm": 0.3434034287929535,
|
1674 |
+
"learning_rate": 5.6510975484659675e-05,
|
1675 |
+
"loss": 0.0399,
|
1676 |
+
"step": 2380
|
1677 |
+
},
|
1678 |
+
{
|
1679 |
+
"epoch": 4.877551020408164,
|
1680 |
+
"grad_norm": 0.4160457253456116,
|
1681 |
+
"learning_rate": 5.617626018911079e-05,
|
1682 |
+
"loss": 0.0494,
|
1683 |
+
"step": 2390
|
1684 |
+
},
|
1685 |
+
{
|
1686 |
+
"epoch": 4.8979591836734695,
|
1687 |
+
"grad_norm": 0.3816176652908325,
|
1688 |
+
"learning_rate": 5.5841263584119594e-05,
|
1689 |
+
"loss": 0.048,
|
1690 |
+
"step": 2400
|
1691 |
+
},
|
1692 |
+
{
|
1693 |
+
"epoch": 4.918367346938775,
|
1694 |
+
"grad_norm": 0.5131191611289978,
|
1695 |
+
"learning_rate": 5.550600092773959e-05,
|
1696 |
+
"loss": 0.0409,
|
1697 |
+
"step": 2410
|
1698 |
+
},
|
1699 |
+
{
|
1700 |
+
"epoch": 4.938775510204081,
|
1701 |
+
"grad_norm": 0.37583550810813904,
|
1702 |
+
"learning_rate": 5.5170487490142006e-05,
|
1703 |
+
"loss": 0.0445,
|
1704 |
+
"step": 2420
|
1705 |
+
},
|
1706 |
+
{
|
1707 |
+
"epoch": 4.959183673469388,
|
1708 |
+
"grad_norm": 0.397798091173172,
|
1709 |
+
"learning_rate": 5.483473855292043e-05,
|
1710 |
+
"loss": 0.04,
|
1711 |
+
"step": 2430
|
1712 |
+
},
|
1713 |
+
{
|
1714 |
+
"epoch": 4.979591836734694,
|
1715 |
+
"grad_norm": 0.5067810416221619,
|
1716 |
+
"learning_rate": 5.449876940839472e-05,
|
1717 |
+
"loss": 0.0476,
|
1718 |
+
"step": 2440
|
1719 |
+
},
|
1720 |
+
{
|
1721 |
+
"epoch": 5.0,
|
1722 |
+
"grad_norm": 0.5477608442306519,
|
1723 |
+
"learning_rate": 5.416259535891447e-05,
|
1724 |
+
"loss": 0.0454,
|
1725 |
+
"step": 2450
|
1726 |
+
},
|
1727 |
+
{
|
1728 |
+
"epoch": 5.020408163265306,
|
1729 |
+
"grad_norm": 0.5660824179649353,
|
1730 |
+
"learning_rate": 5.3826231716162076e-05,
|
1731 |
+
"loss": 0.048,
|
1732 |
+
"step": 2460
|
1733 |
+
},
|
1734 |
+
{
|
1735 |
+
"epoch": 5.040816326530612,
|
1736 |
+
"grad_norm": 0.2889501750469208,
|
1737 |
+
"learning_rate": 5.348969380045531e-05,
|
1738 |
+
"loss": 0.0438,
|
1739 |
+
"step": 2470
|
1740 |
+
},
|
1741 |
+
{
|
1742 |
+
"epoch": 5.061224489795919,
|
1743 |
+
"grad_norm": 0.426714152097702,
|
1744 |
+
"learning_rate": 5.315299694004954e-05,
|
1745 |
+
"loss": 0.0404,
|
1746 |
+
"step": 2480
|
1747 |
+
},
|
1748 |
+
{
|
1749 |
+
"epoch": 5.081632653061225,
|
1750 |
+
"grad_norm": 0.34856680035591125,
|
1751 |
+
"learning_rate": 5.281615647043956e-05,
|
1752 |
+
"loss": 0.0394,
|
1753 |
+
"step": 2490
|
1754 |
+
},
|
1755 |
+
{
|
1756 |
+
"epoch": 5.1020408163265305,
|
1757 |
+
"grad_norm": 0.5608141422271729,
|
1758 |
+
"learning_rate": 5.247918773366112e-05,
|
1759 |
+
"loss": 0.0402,
|
1760 |
+
"step": 2500
|
1761 |
+
},
|
1762 |
+
{
|
1763 |
+
"epoch": 5.122448979591836,
|
1764 |
+
"grad_norm": 0.5281340479850769,
|
1765 |
+
"learning_rate": 5.2142106077592155e-05,
|
1766 |
+
"loss": 0.0397,
|
1767 |
+
"step": 2510
|
1768 |
+
},
|
1769 |
+
{
|
1770 |
+
"epoch": 5.142857142857143,
|
1771 |
+
"grad_norm": 0.48721784353256226,
|
1772 |
+
"learning_rate": 5.18049268552537e-05,
|
1773 |
+
"loss": 0.0435,
|
1774 |
+
"step": 2520
|
1775 |
+
},
|
1776 |
+
{
|
1777 |
+
"epoch": 5.163265306122449,
|
1778 |
+
"grad_norm": 0.4792710542678833,
|
1779 |
+
"learning_rate": 5.1467665424110657e-05,
|
1780 |
+
"loss": 0.0406,
|
1781 |
+
"step": 2530
|
1782 |
+
},
|
1783 |
+
{
|
1784 |
+
"epoch": 5.183673469387755,
|
1785 |
+
"grad_norm": 0.5936989784240723,
|
1786 |
+
"learning_rate": 5.113033714537226e-05,
|
1787 |
+
"loss": 0.0393,
|
1788 |
+
"step": 2540
|
1789 |
+
},
|
1790 |
+
{
|
1791 |
+
"epoch": 5.204081632653061,
|
1792 |
+
"grad_norm": 0.5217839479446411,
|
1793 |
+
"learning_rate": 5.079295738329245e-05,
|
1794 |
+
"loss": 0.042,
|
1795 |
+
"step": 2550
|
1796 |
+
},
|
1797 |
+
{
|
1798 |
+
"epoch": 5.224489795918367,
|
1799 |
+
"grad_norm": 0.4166584610939026,
|
1800 |
+
"learning_rate": 5.0455541504470086e-05,
|
1801 |
+
"loss": 0.0416,
|
1802 |
+
"step": 2560
|
1803 |
+
},
|
1804 |
+
{
|
1805 |
+
"epoch": 5.244897959183674,
|
1806 |
+
"grad_norm": 0.3131585419178009,
|
1807 |
+
"learning_rate": 5.011810487714901e-05,
|
1808 |
+
"loss": 0.0435,
|
1809 |
+
"step": 2570
|
1810 |
+
},
|
1811 |
+
{
|
1812 |
+
"epoch": 5.26530612244898,
|
1813 |
+
"grad_norm": 0.39630407094955444,
|
1814 |
+
"learning_rate": 4.97806628705181e-05,
|
1815 |
+
"loss": 0.0423,
|
1816 |
+
"step": 2580
|
1817 |
+
},
|
1818 |
+
{
|
1819 |
+
"epoch": 5.285714285714286,
|
1820 |
+
"grad_norm": 0.5689300298690796,
|
1821 |
+
"learning_rate": 4.9443230854011255e-05,
|
1822 |
+
"loss": 0.0443,
|
1823 |
+
"step": 2590
|
1824 |
+
},
|
1825 |
+
{
|
1826 |
+
"epoch": 5.3061224489795915,
|
1827 |
+
"grad_norm": 0.36981523036956787,
|
1828 |
+
"learning_rate": 4.910582419660735e-05,
|
1829 |
+
"loss": 0.0381,
|
1830 |
+
"step": 2600
|
1831 |
+
},
|
1832 |
+
{
|
1833 |
+
"epoch": 5.326530612244898,
|
1834 |
+
"grad_norm": 0.5300806164741516,
|
1835 |
+
"learning_rate": 4.876845826613026e-05,
|
1836 |
+
"loss": 0.0385,
|
1837 |
+
"step": 2610
|
1838 |
+
},
|
1839 |
+
{
|
1840 |
+
"epoch": 5.346938775510204,
|
1841 |
+
"grad_norm": 0.37739285826683044,
|
1842 |
+
"learning_rate": 4.843114842854881e-05,
|
1843 |
+
"loss": 0.0389,
|
1844 |
+
"step": 2620
|
1845 |
+
},
|
1846 |
+
{
|
1847 |
+
"epoch": 5.36734693877551,
|
1848 |
+
"grad_norm": 0.288785457611084,
|
1849 |
+
"learning_rate": 4.8093910047277e-05,
|
1850 |
+
"loss": 0.0397,
|
1851 |
+
"step": 2630
|
1852 |
+
},
|
1853 |
+
{
|
1854 |
+
"epoch": 5.387755102040816,
|
1855 |
+
"grad_norm": 0.37196066975593567,
|
1856 |
+
"learning_rate": 4.775675848247427e-05,
|
1857 |
+
"loss": 0.0365,
|
1858 |
+
"step": 2640
|
1859 |
+
},
|
1860 |
+
{
|
1861 |
+
"epoch": 5.408163265306122,
|
1862 |
+
"grad_norm": 0.4045736789703369,
|
1863 |
+
"learning_rate": 4.7419709090345734e-05,
|
1864 |
+
"loss": 0.0357,
|
1865 |
+
"step": 2650
|
1866 |
+
},
|
1867 |
+
{
|
1868 |
+
"epoch": 5.428571428571429,
|
1869 |
+
"grad_norm": 0.3250954747200012,
|
1870 |
+
"learning_rate": 4.7082777222442954e-05,
|
1871 |
+
"loss": 0.041,
|
1872 |
+
"step": 2660
|
1873 |
+
},
|
1874 |
+
{
|
1875 |
+
"epoch": 5.448979591836735,
|
1876 |
+
"grad_norm": 0.4290127754211426,
|
1877 |
+
"learning_rate": 4.6745978224964574e-05,
|
1878 |
+
"loss": 0.0419,
|
1879 |
+
"step": 2670
|
1880 |
+
},
|
1881 |
+
{
|
1882 |
+
"epoch": 5.469387755102041,
|
1883 |
+
"grad_norm": 0.40228599309921265,
|
1884 |
+
"learning_rate": 4.640932743805745e-05,
|
1885 |
+
"loss": 0.0406,
|
1886 |
+
"step": 2680
|
1887 |
+
},
|
1888 |
+
{
|
1889 |
+
"epoch": 5.489795918367347,
|
1890 |
+
"grad_norm": 0.42489856481552124,
|
1891 |
+
"learning_rate": 4.607284019511784e-05,
|
1892 |
+
"loss": 0.0353,
|
1893 |
+
"step": 2690
|
1894 |
+
},
|
1895 |
+
{
|
1896 |
+
"epoch": 5.510204081632653,
|
1897 |
+
"grad_norm": 0.6084537506103516,
|
1898 |
+
"learning_rate": 4.5736531822093136e-05,
|
1899 |
+
"loss": 0.0419,
|
1900 |
+
"step": 2700
|
1901 |
+
},
|
1902 |
+
{
|
1903 |
+
"epoch": 5.530612244897959,
|
1904 |
+
"grad_norm": 0.4374919533729553,
|
1905 |
+
"learning_rate": 4.540041763678378e-05,
|
1906 |
+
"loss": 0.0377,
|
1907 |
+
"step": 2710
|
1908 |
+
},
|
1909 |
+
{
|
1910 |
+
"epoch": 5.551020408163265,
|
1911 |
+
"grad_norm": 0.4828069508075714,
|
1912 |
+
"learning_rate": 4.5064512948145474e-05,
|
1913 |
+
"loss": 0.0408,
|
1914 |
+
"step": 2720
|
1915 |
+
},
|
1916 |
+
{
|
1917 |
+
"epoch": 5.571428571428571,
|
1918 |
+
"grad_norm": 0.4128870964050293,
|
1919 |
+
"learning_rate": 4.472883305559209e-05,
|
1920 |
+
"loss": 0.043,
|
1921 |
+
"step": 2730
|
1922 |
+
},
|
1923 |
+
{
|
1924 |
+
"epoch": 5.591836734693878,
|
1925 |
+
"grad_norm": 0.47663477063179016,
|
1926 |
+
"learning_rate": 4.439339324829872e-05,
|
1927 |
+
"loss": 0.0429,
|
1928 |
+
"step": 2740
|
1929 |
+
},
|
1930 |
+
{
|
1931 |
+
"epoch": 5.612244897959184,
|
1932 |
+
"grad_norm": 0.46140581369400024,
|
1933 |
+
"learning_rate": 4.405820880450529e-05,
|
1934 |
+
"loss": 0.0392,
|
1935 |
+
"step": 2750
|
1936 |
+
},
|
1937 |
+
{
|
1938 |
+
"epoch": 5.63265306122449,
|
1939 |
+
"grad_norm": 0.4380878806114197,
|
1940 |
+
"learning_rate": 4.372329499082072e-05,
|
1941 |
+
"loss": 0.042,
|
1942 |
+
"step": 2760
|
1943 |
+
},
|
1944 |
+
{
|
1945 |
+
"epoch": 5.653061224489796,
|
1946 |
+
"grad_norm": 0.34445661306381226,
|
1947 |
+
"learning_rate": 4.33886670615276e-05,
|
1948 |
+
"loss": 0.0449,
|
1949 |
+
"step": 2770
|
1950 |
+
},
|
1951 |
+
{
|
1952 |
+
"epoch": 5.673469387755102,
|
1953 |
+
"grad_norm": 0.4100647270679474,
|
1954 |
+
"learning_rate": 4.305434025788734e-05,
|
1955 |
+
"loss": 0.0401,
|
1956 |
+
"step": 2780
|
1957 |
+
},
|
1958 |
+
{
|
1959 |
+
"epoch": 5.6938775510204085,
|
1960 |
+
"grad_norm": 0.31862297654151917,
|
1961 |
+
"learning_rate": 4.272032980744603e-05,
|
1962 |
+
"loss": 0.0344,
|
1963 |
+
"step": 2790
|
1964 |
+
},
|
1965 |
+
{
|
1966 |
+
"epoch": 5.714285714285714,
|
1967 |
+
"grad_norm": 0.5393098592758179,
|
1968 |
+
"learning_rate": 4.238665092334084e-05,
|
1969 |
+
"loss": 0.039,
|
1970 |
+
"step": 2800
|
1971 |
+
},
|
1972 |
+
{
|
1973 |
+
"epoch": 5.73469387755102,
|
1974 |
+
"grad_norm": 0.42914947867393494,
|
1975 |
+
"learning_rate": 4.2053318803607156e-05,
|
1976 |
+
"loss": 0.0425,
|
1977 |
+
"step": 2810
|
1978 |
+
},
|
1979 |
+
{
|
1980 |
+
"epoch": 5.755102040816326,
|
1981 |
+
"grad_norm": 0.3658852279186249,
|
1982 |
+
"learning_rate": 4.172034863048631e-05,
|
1983 |
+
"loss": 0.029,
|
1984 |
+
"step": 2820
|
1985 |
+
},
|
1986 |
+
{
|
1987 |
+
"epoch": 5.775510204081632,
|
1988 |
+
"grad_norm": 0.3325256109237671,
|
1989 |
+
"learning_rate": 4.138775556973406e-05,
|
1990 |
+
"loss": 0.0407,
|
1991 |
+
"step": 2830
|
1992 |
+
},
|
1993 |
+
{
|
1994 |
+
"epoch": 5.795918367346939,
|
1995 |
+
"grad_norm": 0.3700428903102875,
|
1996 |
+
"learning_rate": 4.105555476992992e-05,
|
1997 |
+
"loss": 0.0393,
|
1998 |
+
"step": 2840
|
1999 |
+
},
|
2000 |
+
{
|
2001 |
+
"epoch": 5.816326530612245,
|
2002 |
+
"grad_norm": 0.4134577810764313,
|
2003 |
+
"learning_rate": 4.0723761361787116e-05,
|
2004 |
+
"loss": 0.0388,
|
2005 |
+
"step": 2850
|
2006 |
+
},
|
2007 |
+
{
|
2008 |
+
"epoch": 5.836734693877551,
|
2009 |
+
"grad_norm": 0.4173426032066345,
|
2010 |
+
"learning_rate": 4.039239045746342e-05,
|
2011 |
+
"loss": 0.0404,
|
2012 |
+
"step": 2860
|
2013 |
+
},
|
2014 |
+
{
|
2015 |
+
"epoch": 5.857142857142857,
|
2016 |
+
"grad_norm": 0.3706321716308594,
|
2017 |
+
"learning_rate": 4.006145714987293e-05,
|
2018 |
+
"loss": 0.0367,
|
2019 |
+
"step": 2870
|
2020 |
+
},
|
2021 |
+
{
|
2022 |
+
"epoch": 5.877551020408164,
|
2023 |
+
"grad_norm": 0.34077945351600647,
|
2024 |
+
"learning_rate": 3.973097651199853e-05,
|
2025 |
+
"loss": 0.0349,
|
2026 |
+
"step": 2880
|
2027 |
+
},
|
2028 |
+
{
|
2029 |
+
"epoch": 5.8979591836734695,
|
2030 |
+
"grad_norm": 0.48339715600013733,
|
2031 |
+
"learning_rate": 3.9400963596205434e-05,
|
2032 |
+
"loss": 0.0359,
|
2033 |
+
"step": 2890
|
2034 |
+
},
|
2035 |
+
{
|
2036 |
+
"epoch": 5.918367346938775,
|
2037 |
+
"grad_norm": 0.47534608840942383,
|
2038 |
+
"learning_rate": 3.9071433433555525e-05,
|
2039 |
+
"loss": 0.04,
|
2040 |
+
"step": 2900
|
2041 |
+
},
|
2042 |
+
{
|
2043 |
+
"epoch": 5.938775510204081,
|
2044 |
+
"grad_norm": 0.4137221872806549,
|
2045 |
+
"learning_rate": 3.874240103312282e-05,
|
2046 |
+
"loss": 0.0489,
|
2047 |
+
"step": 2910
|
2048 |
+
},
|
2049 |
+
{
|
2050 |
+
"epoch": 5.959183673469388,
|
2051 |
+
"grad_norm": 0.3204796314239502,
|
2052 |
+
"learning_rate": 3.841388138130984e-05,
|
2053 |
+
"loss": 0.0341,
|
2054 |
+
"step": 2920
|
2055 |
+
},
|
2056 |
+
{
|
2057 |
+
"epoch": 5.979591836734694,
|
2058 |
+
"grad_norm": 0.5732039213180542,
|
2059 |
+
"learning_rate": 3.808588944116491e-05,
|
2060 |
+
"loss": 0.0342,
|
2061 |
+
"step": 2930
|
2062 |
+
},
|
2063 |
+
{
|
2064 |
+
"epoch": 6.0,
|
2065 |
+
"grad_norm": 0.7128877639770508,
|
2066 |
+
"learning_rate": 3.775844015170083e-05,
|
2067 |
+
"loss": 0.0374,
|
2068 |
+
"step": 2940
|
2069 |
+
},
|
2070 |
+
{
|
2071 |
+
"epoch": 6.020408163265306,
|
2072 |
+
"grad_norm": 0.33346930146217346,
|
2073 |
+
"learning_rate": 3.7431548427214295e-05,
|
2074 |
+
"loss": 0.0358,
|
2075 |
+
"step": 2950
|
2076 |
+
},
|
2077 |
+
{
|
2078 |
+
"epoch": 6.040816326530612,
|
2079 |
+
"grad_norm": 0.41604581475257874,
|
2080 |
+
"learning_rate": 3.710522915660666e-05,
|
2081 |
+
"loss": 0.0456,
|
2082 |
+
"step": 2960
|
2083 |
+
},
|
2084 |
+
{
|
2085 |
+
"epoch": 6.061224489795919,
|
2086 |
+
"grad_norm": 0.344349205493927,
|
2087 |
+
"learning_rate": 3.677949720270578e-05,
|
2088 |
+
"loss": 0.0384,
|
2089 |
+
"step": 2970
|
2090 |
+
},
|
2091 |
+
{
|
2092 |
+
"epoch": 6.081632653061225,
|
2093 |
+
"grad_norm": 1.0503120422363281,
|
2094 |
+
"learning_rate": 3.645436740158908e-05,
|
2095 |
+
"loss": 0.0383,
|
2096 |
+
"step": 2980
|
2097 |
+
},
|
2098 |
+
{
|
2099 |
+
"epoch": 6.1020408163265305,
|
2100 |
+
"grad_norm": 0.35643085837364197,
|
2101 |
+
"learning_rate": 3.612985456190778e-05,
|
2102 |
+
"loss": 0.0312,
|
2103 |
+
"step": 2990
|
2104 |
+
},
|
2105 |
+
{
|
2106 |
+
"epoch": 6.122448979591836,
|
2107 |
+
"grad_norm": 0.2821815609931946,
|
2108 |
+
"learning_rate": 3.580597346421243e-05,
|
2109 |
+
"loss": 0.0342,
|
2110 |
+
"step": 3000
|
2111 |
+
},
|
2112 |
+
{
|
2113 |
+
"epoch": 6.142857142857143,
|
2114 |
+
"grad_norm": 0.3137131631374359,
|
2115 |
+
"learning_rate": 3.54827388602797e-05,
|
2116 |
+
"loss": 0.038,
|
2117 |
+
"step": 3010
|
2118 |
+
},
|
2119 |
+
{
|
2120 |
+
"epoch": 6.163265306122449,
|
2121 |
+
"grad_norm": 0.44174015522003174,
|
2122 |
+
"learning_rate": 3.5160165472440473e-05,
|
2123 |
+
"loss": 0.0379,
|
2124 |
+
"step": 3020
|
2125 |
+
},
|
2126 |
+
{
|
2127 |
+
"epoch": 6.183673469387755,
|
2128 |
+
"grad_norm": 0.3963649868965149,
|
2129 |
+
"learning_rate": 3.4838267992909325e-05,
|
2130 |
+
"loss": 0.035,
|
2131 |
+
"step": 3030
|
2132 |
+
},
|
2133 |
+
{
|
2134 |
+
"epoch": 6.204081632653061,
|
2135 |
+
"grad_norm": 0.3708857595920563,
|
2136 |
+
"learning_rate": 3.451706108311525e-05,
|
2137 |
+
"loss": 0.0342,
|
2138 |
+
"step": 3040
|
2139 |
+
},
|
2140 |
+
{
|
2141 |
+
"epoch": 6.224489795918367,
|
2142 |
+
"grad_norm": 0.3640486001968384,
|
2143 |
+
"learning_rate": 3.4196559373034006e-05,
|
2144 |
+
"loss": 0.0327,
|
2145 |
+
"step": 3050
|
2146 |
+
},
|
2147 |
+
{
|
2148 |
+
"epoch": 6.244897959183674,
|
2149 |
+
"grad_norm": 0.5147858262062073,
|
2150 |
+
"learning_rate": 3.387677746052165e-05,
|
2151 |
+
"loss": 0.0453,
|
2152 |
+
"step": 3060
|
2153 |
+
},
|
2154 |
+
{
|
2155 |
+
"epoch": 6.26530612244898,
|
2156 |
+
"grad_norm": 0.5431321859359741,
|
2157 |
+
"learning_rate": 3.3557729910649734e-05,
|
2158 |
+
"loss": 0.0376,
|
2159 |
+
"step": 3070
|
2160 |
+
},
|
2161 |
+
{
|
2162 |
+
"epoch": 6.285714285714286,
|
2163 |
+
"grad_norm": 0.41198182106018066,
|
2164 |
+
"learning_rate": 3.323943125504184e-05,
|
2165 |
+
"loss": 0.0354,
|
2166 |
+
"step": 3080
|
2167 |
+
},
|
2168 |
+
{
|
2169 |
+
"epoch": 6.3061224489795915,
|
2170 |
+
"grad_norm": 0.43326935172080994,
|
2171 |
+
"learning_rate": 3.292189599121176e-05,
|
2172 |
+
"loss": 0.0327,
|
2173 |
+
"step": 3090
|
2174 |
+
},
|
2175 |
+
{
|
2176 |
+
"epoch": 6.326530612244898,
|
2177 |
+
"grad_norm": 0.38192853331565857,
|
2178 |
+
"learning_rate": 3.260513858190319e-05,
|
2179 |
+
"loss": 0.0352,
|
2180 |
+
"step": 3100
|
2181 |
+
},
|
2182 |
+
{
|
2183 |
+
"epoch": 6.346938775510204,
|
2184 |
+
"grad_norm": 0.41123682260513306,
|
2185 |
+
"learning_rate": 3.228917345443092e-05,
|
2186 |
+
"loss": 0.0377,
|
2187 |
+
"step": 3110
|
2188 |
+
},
|
2189 |
+
{
|
2190 |
+
"epoch": 6.36734693877551,
|
2191 |
+
"grad_norm": 0.5325567126274109,
|
2192 |
+
"learning_rate": 3.19740150000238e-05,
|
2193 |
+
"loss": 0.0356,
|
2194 |
+
"step": 3120
|
2195 |
+
},
|
2196 |
+
{
|
2197 |
+
"epoch": 6.387755102040816,
|
2198 |
+
"grad_norm": 0.3802492916584015,
|
2199 |
+
"learning_rate": 3.165967757316925e-05,
|
2200 |
+
"loss": 0.0351,
|
2201 |
+
"step": 3130
|
2202 |
+
},
|
2203 |
+
{
|
2204 |
+
"epoch": 6.408163265306122,
|
2205 |
+
"grad_norm": 0.26050621271133423,
|
2206 |
+
"learning_rate": 3.134617549095941e-05,
|
2207 |
+
"loss": 0.0363,
|
2208 |
+
"step": 3140
|
2209 |
+
},
|
2210 |
+
{
|
2211 |
+
"epoch": 6.428571428571429,
|
2212 |
+
"grad_norm": 0.3772394359111786,
|
2213 |
+
"learning_rate": 3.103352303243905e-05,
|
2214 |
+
"loss": 0.0353,
|
2215 |
+
"step": 3150
|
2216 |
+
},
|
2217 |
+
{
|
2218 |
+
"epoch": 6.448979591836735,
|
2219 |
+
"grad_norm": 0.36528316140174866,
|
2220 |
+
"learning_rate": 3.072173443795527e-05,
|
2221 |
+
"loss": 0.0312,
|
2222 |
+
"step": 3160
|
2223 |
+
},
|
2224 |
+
{
|
2225 |
+
"epoch": 6.469387755102041,
|
2226 |
+
"grad_norm": 0.41683340072631836,
|
2227 |
+
"learning_rate": 3.0410823908508816e-05,
|
2228 |
+
"loss": 0.0447,
|
2229 |
+
"step": 3170
|
2230 |
+
},
|
2231 |
+
{
|
2232 |
+
"epoch": 6.489795918367347,
|
2233 |
+
"grad_norm": 0.41402310132980347,
|
2234 |
+
"learning_rate": 3.0100805605107312e-05,
|
2235 |
+
"loss": 0.0311,
|
2236 |
+
"step": 3180
|
2237 |
+
},
|
2238 |
+
{
|
2239 |
+
"epoch": 6.510204081632653,
|
2240 |
+
"grad_norm": 0.4392901062965393,
|
2241 |
+
"learning_rate": 2.979169364812028e-05,
|
2242 |
+
"loss": 0.0325,
|
2243 |
+
"step": 3190
|
2244 |
+
},
|
2245 |
+
{
|
2246 |
+
"epoch": 6.530612244897959,
|
2247 |
+
"grad_norm": 0.3072589933872223,
|
2248 |
+
"learning_rate": 2.948350211663594e-05,
|
2249 |
+
"loss": 0.0385,
|
2250 |
+
"step": 3200
|
2251 |
+
},
|
2252 |
+
{
|
2253 |
+
"epoch": 6.551020408163265,
|
2254 |
+
"grad_norm": 0.3144962787628174,
|
2255 |
+
"learning_rate": 2.917624504782006e-05,
|
2256 |
+
"loss": 0.0298,
|
2257 |
+
"step": 3210
|
2258 |
+
},
|
2259 |
+
{
|
2260 |
+
"epoch": 6.571428571428571,
|
2261 |
+
"grad_norm": 0.33626478910446167,
|
2262 |
+
"learning_rate": 2.886993643627644e-05,
|
2263 |
+
"loss": 0.033,
|
2264 |
+
"step": 3220
|
2265 |
+
},
|
2266 |
+
{
|
2267 |
+
"epoch": 6.591836734693878,
|
2268 |
+
"grad_norm": 0.4536080062389374,
|
2269 |
+
"learning_rate": 2.8564590233409693e-05,
|
2270 |
+
"loss": 0.0313,
|
2271 |
+
"step": 3230
|
2272 |
+
},
|
2273 |
+
{
|
2274 |
+
"epoch": 6.612244897959184,
|
2275 |
+
"grad_norm": 0.25697794556617737,
|
2276 |
+
"learning_rate": 2.826022034678969e-05,
|
2277 |
+
"loss": 0.0286,
|
2278 |
+
"step": 3240
|
2279 |
+
},
|
2280 |
+
{
|
2281 |
+
"epoch": 6.63265306122449,
|
2282 |
+
"grad_norm": 0.295259952545166,
|
2283 |
+
"learning_rate": 2.7956840639518077e-05,
|
2284 |
+
"loss": 0.0239,
|
2285 |
+
"step": 3250
|
2286 |
+
},
|
2287 |
+
{
|
2288 |
+
"epoch": 6.653061224489796,
|
2289 |
+
"grad_norm": 0.3460862934589386,
|
2290 |
+
"learning_rate": 2.7654464929596975e-05,
|
2291 |
+
"loss": 0.037,
|
2292 |
+
"step": 3260
|
2293 |
+
},
|
2294 |
+
{
|
2295 |
+
"epoch": 6.673469387755102,
|
2296 |
+
"grad_norm": 0.3951895534992218,
|
2297 |
+
"learning_rate": 2.7353106989299528e-05,
|
2298 |
+
"loss": 0.0306,
|
2299 |
+
"step": 3270
|
2300 |
+
},
|
2301 |
+
{
|
2302 |
+
"epoch": 6.6938775510204085,
|
2303 |
+
"grad_norm": 0.268476665019989,
|
2304 |
+
"learning_rate": 2.705278054454265e-05,
|
2305 |
+
"loss": 0.0299,
|
2306 |
+
"step": 3280
|
2307 |
+
},
|
2308 |
+
{
|
2309 |
+
"epoch": 6.714285714285714,
|
2310 |
+
"grad_norm": 0.37184420228004456,
|
2311 |
+
"learning_rate": 2.6753499274261795e-05,
|
2312 |
+
"loss": 0.0326,
|
2313 |
+
"step": 3290
|
2314 |
+
},
|
2315 |
+
{
|
2316 |
+
"epoch": 6.73469387755102,
|
2317 |
+
"grad_norm": 0.33349689841270447,
|
2318 |
+
"learning_rate": 2.6455276809788023e-05,
|
2319 |
+
"loss": 0.0269,
|
2320 |
+
"step": 3300
|
2321 |
+
},
|
2322 |
+
{
|
2323 |
+
"epoch": 6.755102040816326,
|
2324 |
+
"grad_norm": 0.30202341079711914,
|
2325 |
+
"learning_rate": 2.6158126734227072e-05,
|
2326 |
+
"loss": 0.0291,
|
2327 |
+
"step": 3310
|
2328 |
+
},
|
2329 |
+
{
|
2330 |
+
"epoch": 6.775510204081632,
|
2331 |
+
"grad_norm": 0.3205631971359253,
|
2332 |
+
"learning_rate": 2.5862062581840674e-05,
|
2333 |
+
"loss": 0.0308,
|
2334 |
+
"step": 3320
|
2335 |
+
},
|
2336 |
+
{
|
2337 |
+
"epoch": 6.795918367346939,
|
2338 |
+
"grad_norm": 0.2775018513202667,
|
2339 |
+
"learning_rate": 2.5567097837430155e-05,
|
2340 |
+
"loss": 0.0287,
|
2341 |
+
"step": 3330
|
2342 |
+
},
|
2343 |
+
{
|
2344 |
+
"epoch": 6.816326530612245,
|
2345 |
+
"grad_norm": 0.3472177982330322,
|
2346 |
+
"learning_rate": 2.5273245935722227e-05,
|
2347 |
+
"loss": 0.0356,
|
2348 |
+
"step": 3340
|
2349 |
+
},
|
2350 |
+
{
|
2351 |
+
"epoch": 6.836734693877551,
|
2352 |
+
"grad_norm": 0.40243059396743774,
|
2353 |
+
"learning_rate": 2.498052026075711e-05,
|
2354 |
+
"loss": 0.0327,
|
2355 |
+
"step": 3350
|
2356 |
+
},
|
2357 |
+
{
|
2358 |
+
"epoch": 6.857142857142857,
|
2359 |
+
"grad_norm": 0.34365713596343994,
|
2360 |
+
"learning_rate": 2.4688934145278825e-05,
|
2361 |
+
"loss": 0.0325,
|
2362 |
+
"step": 3360
|
2363 |
+
},
|
2364 |
+
{
|
2365 |
+
"epoch": 6.877551020408164,
|
2366 |
+
"grad_norm": 0.6802788972854614,
|
2367 |
+
"learning_rate": 2.439850087012806e-05,
|
2368 |
+
"loss": 0.0382,
|
2369 |
+
"step": 3370
|
2370 |
+
},
|
2371 |
+
{
|
2372 |
+
"epoch": 6.8979591836734695,
|
2373 |
+
"grad_norm": 0.3928966522216797,
|
2374 |
+
"learning_rate": 2.4109233663637215e-05,
|
2375 |
+
"loss": 0.0289,
|
2376 |
+
"step": 3380
|
2377 |
+
},
|
2378 |
+
{
|
2379 |
+
"epoch": 6.918367346938775,
|
2380 |
+
"grad_norm": 0.35136857628822327,
|
2381 |
+
"learning_rate": 2.3821145701027836e-05,
|
2382 |
+
"loss": 0.0375,
|
2383 |
+
"step": 3390
|
2384 |
+
},
|
2385 |
+
{
|
2386 |
+
"epoch": 6.938775510204081,
|
2387 |
+
"grad_norm": 0.42939406633377075,
|
2388 |
+
"learning_rate": 2.353425010381063e-05,
|
2389 |
+
"loss": 0.03,
|
2390 |
+
"step": 3400
|
2391 |
+
},
|
2392 |
+
{
|
2393 |
+
"epoch": 6.959183673469388,
|
2394 |
+
"grad_norm": 0.2758863866329193,
|
2395 |
+
"learning_rate": 2.3248559939187748e-05,
|
2396 |
+
"loss": 0.0339,
|
2397 |
+
"step": 3410
|
2398 |
+
},
|
2399 |
+
{
|
2400 |
+
"epoch": 6.979591836734694,
|
2401 |
+
"grad_norm": 0.4242722690105438,
|
2402 |
+
"learning_rate": 2.2964088219457664e-05,
|
2403 |
+
"loss": 0.0317,
|
2404 |
+
"step": 3420
|
2405 |
+
},
|
2406 |
+
{
|
2407 |
+
"epoch": 7.0,
|
2408 |
+
"grad_norm": 0.31365418434143066,
|
2409 |
+
"learning_rate": 2.2680847901422403e-05,
|
2410 |
+
"loss": 0.0316,
|
2411 |
+
"step": 3430
|
2412 |
+
},
|
2413 |
+
{
|
2414 |
+
"epoch": 7.020408163265306,
|
2415 |
+
"grad_norm": 0.3401188254356384,
|
2416 |
+
"learning_rate": 2.239885188579755e-05,
|
2417 |
+
"loss": 0.025,
|
2418 |
+
"step": 3440
|
2419 |
+
},
|
2420 |
+
{
|
2421 |
+
"epoch": 7.040816326530612,
|
2422 |
+
"grad_norm": 0.4100947678089142,
|
2423 |
+
"learning_rate": 2.211811301662457e-05,
|
2424 |
+
"loss": 0.0273,
|
2425 |
+
"step": 3450
|
2426 |
+
},
|
2427 |
+
{
|
2428 |
+
"epoch": 7.061224489795919,
|
2429 |
+
"grad_norm": 0.4056410491466522,
|
2430 |
+
"learning_rate": 2.183864408068577e-05,
|
2431 |
+
"loss": 0.0357,
|
2432 |
+
"step": 3460
|
2433 |
+
},
|
2434 |
+
{
|
2435 |
+
"epoch": 7.081632653061225,
|
2436 |
+
"grad_norm": 0.42886030673980713,
|
2437 |
+
"learning_rate": 2.1560457806921997e-05,
|
2438 |
+
"loss": 0.0299,
|
2439 |
+
"step": 3470
|
2440 |
+
},
|
2441 |
+
{
|
2442 |
+
"epoch": 7.1020408163265305,
|
2443 |
+
"grad_norm": 0.4000697731971741,
|
2444 |
+
"learning_rate": 2.128356686585282e-05,
|
2445 |
+
"loss": 0.0296,
|
2446 |
+
"step": 3480
|
2447 |
+
},
|
2448 |
+
{
|
2449 |
+
"epoch": 7.122448979591836,
|
2450 |
+
"grad_norm": 0.29271119832992554,
|
2451 |
+
"learning_rate": 2.1007983868999405e-05,
|
2452 |
+
"loss": 0.0294,
|
2453 |
+
"step": 3490
|
2454 |
+
},
|
2455 |
+
{
|
2456 |
+
"epoch": 7.142857142857143,
|
2457 |
+
"grad_norm": 0.3026485741138458,
|
2458 |
+
"learning_rate": 2.0733721368310144e-05,
|
2459 |
+
"loss": 0.0349,
|
2460 |
+
"step": 3500
|
2461 |
+
},
|
2462 |
+
{
|
2463 |
+
"epoch": 7.163265306122449,
|
2464 |
+
"grad_norm": 0.5303738713264465,
|
2465 |
+
"learning_rate": 2.0460791855588957e-05,
|
2466 |
+
"loss": 0.0292,
|
2467 |
+
"step": 3510
|
2468 |
+
},
|
2469 |
+
{
|
2470 |
+
"epoch": 7.183673469387755,
|
2471 |
+
"grad_norm": 0.47380366921424866,
|
2472 |
+
"learning_rate": 2.018920776192626e-05,
|
2473 |
+
"loss": 0.0275,
|
2474 |
+
"step": 3520
|
2475 |
+
},
|
2476 |
+
{
|
2477 |
+
"epoch": 7.204081632653061,
|
2478 |
+
"grad_norm": 0.4064404368400574,
|
2479 |
+
"learning_rate": 1.991898145713287e-05,
|
2480 |
+
"loss": 0.0229,
|
2481 |
+
"step": 3530
|
2482 |
+
},
|
2483 |
+
{
|
2484 |
+
"epoch": 7.224489795918367,
|
2485 |
+
"grad_norm": 0.38069799542427063,
|
2486 |
+
"learning_rate": 1.9650125249176475e-05,
|
2487 |
+
"loss": 0.0291,
|
2488 |
+
"step": 3540
|
2489 |
+
},
|
2490 |
+
{
|
2491 |
+
"epoch": 7.244897959183674,
|
2492 |
+
"grad_norm": 0.3179313540458679,
|
2493 |
+
"learning_rate": 1.938265138362118e-05,
|
2494 |
+
"loss": 0.0315,
|
2495 |
+
"step": 3550
|
2496 |
+
},
|
2497 |
+
{
|
2498 |
+
"epoch": 7.26530612244898,
|
2499 |
+
"grad_norm": 0.3878839612007141,
|
2500 |
+
"learning_rate": 1.9116572043069687e-05,
|
2501 |
+
"loss": 0.0313,
|
2502 |
+
"step": 3560
|
2503 |
+
},
|
2504 |
+
{
|
2505 |
+
"epoch": 7.285714285714286,
|
2506 |
+
"grad_norm": 0.32467424869537354,
|
2507 |
+
"learning_rate": 1.885189934660836e-05,
|
2508 |
+
"loss": 0.0285,
|
2509 |
+
"step": 3570
|
2510 |
+
},
|
2511 |
+
{
|
2512 |
+
"epoch": 7.3061224489795915,
|
2513 |
+
"grad_norm": 0.3442543148994446,
|
2514 |
+
"learning_rate": 1.8588645349255364e-05,
|
2515 |
+
"loss": 0.0305,
|
2516 |
+
"step": 3580
|
2517 |
+
},
|
2518 |
+
{
|
2519 |
+
"epoch": 7.326530612244898,
|
2520 |
+
"grad_norm": 0.44325685501098633,
|
2521 |
+
"learning_rate": 1.8326822041411524e-05,
|
2522 |
+
"loss": 0.0275,
|
2523 |
+
"step": 3590
|
2524 |
+
},
|
2525 |
+
{
|
2526 |
+
"epoch": 7.346938775510204,
|
2527 |
+
"grad_norm": 0.3210112750530243,
|
2528 |
+
"learning_rate": 1.806644134831415e-05,
|
2529 |
+
"loss": 0.0262,
|
2530 |
+
"step": 3600
|
2531 |
+
},
|
2532 |
+
{
|
2533 |
+
"epoch": 7.36734693877551,
|
2534 |
+
"grad_norm": 0.31204965710639954,
|
2535 |
+
"learning_rate": 1.7807515129494014e-05,
|
2536 |
+
"loss": 0.0293,
|
2537 |
+
"step": 3610
|
2538 |
+
},
|
2539 |
+
{
|
2540 |
+
"epoch": 7.387755102040816,
|
2541 |
+
"grad_norm": 0.33807334303855896,
|
2542 |
+
"learning_rate": 1.7550055178235058e-05,
|
2543 |
+
"loss": 0.0299,
|
2544 |
+
"step": 3620
|
2545 |
+
},
|
2546 |
+
{
|
2547 |
+
"epoch": 7.408163265306122,
|
2548 |
+
"grad_norm": 0.3368700444698334,
|
2549 |
+
"learning_rate": 1.7294073221037344e-05,
|
2550 |
+
"loss": 0.0321,
|
2551 |
+
"step": 3630
|
2552 |
+
},
|
2553 |
+
{
|
2554 |
+
"epoch": 7.428571428571429,
|
2555 |
+
"grad_norm": 0.401540607213974,
|
2556 |
+
"learning_rate": 1.703958091708282e-05,
|
2557 |
+
"loss": 0.0273,
|
2558 |
+
"step": 3640
|
2559 |
+
},
|
2560 |
+
{
|
2561 |
+
"epoch": 7.448979591836735,
|
2562 |
+
"grad_norm": 0.28700941801071167,
|
2563 |
+
"learning_rate": 1.678658985770445e-05,
|
2564 |
+
"loss": 0.0304,
|
2565 |
+
"step": 3650
|
2566 |
+
},
|
2567 |
+
{
|
2568 |
+
"epoch": 7.469387755102041,
|
2569 |
+
"grad_norm": 0.3458382487297058,
|
2570 |
+
"learning_rate": 1.6535111565858135e-05,
|
2571 |
+
"loss": 0.0286,
|
2572 |
+
"step": 3660
|
2573 |
+
},
|
2574 |
+
{
|
2575 |
+
"epoch": 7.489795918367347,
|
2576 |
+
"grad_norm": 0.4366629123687744,
|
2577 |
+
"learning_rate": 1.6285157495597904e-05,
|
2578 |
+
"loss": 0.0301,
|
2579 |
+
"step": 3670
|
2580 |
+
},
|
2581 |
+
{
|
2582 |
+
"epoch": 7.510204081632653,
|
2583 |
+
"grad_norm": 0.37988075613975525,
|
2584 |
+
"learning_rate": 1.6036739031554266e-05,
|
2585 |
+
"loss": 0.0341,
|
2586 |
+
"step": 3680
|
2587 |
+
},
|
2588 |
+
{
|
2589 |
+
"epoch": 7.530612244897959,
|
2590 |
+
"grad_norm": 0.3958030343055725,
|
2591 |
+
"learning_rate": 1.5789867488415632e-05,
|
2592 |
+
"loss": 0.0351,
|
2593 |
+
"step": 3690
|
2594 |
+
},
|
2595 |
+
{
|
2596 |
+
"epoch": 7.551020408163265,
|
2597 |
+
"grad_norm": 0.30942267179489136,
|
2598 |
+
"learning_rate": 1.5544554110412985e-05,
|
2599 |
+
"loss": 0.0334,
|
2600 |
+
"step": 3700
|
2601 |
+
},
|
2602 |
+
{
|
2603 |
+
"epoch": 7.571428571428571,
|
2604 |
+
"grad_norm": 0.4118305444717407,
|
2605 |
+
"learning_rate": 1.5300810070807697e-05,
|
2606 |
+
"loss": 0.0267,
|
2607 |
+
"step": 3710
|
2608 |
+
},
|
2609 |
+
{
|
2610 |
+
"epoch": 7.591836734693878,
|
2611 |
+
"grad_norm": 0.3243233859539032,
|
2612 |
+
"learning_rate": 1.5058646471382698e-05,
|
2613 |
+
"loss": 0.0374,
|
2614 |
+
"step": 3720
|
2615 |
+
},
|
2616 |
+
{
|
2617 |
+
"epoch": 7.612244897959184,
|
2618 |
+
"grad_norm": 0.32483312487602234,
|
2619 |
+
"learning_rate": 1.4818074341936782e-05,
|
2620 |
+
"loss": 0.0264,
|
2621 |
+
"step": 3730
|
2622 |
+
},
|
2623 |
+
{
|
2624 |
+
"epoch": 7.63265306122449,
|
2625 |
+
"grad_norm": 0.2870495915412903,
|
2626 |
+
"learning_rate": 1.4579104639782188e-05,
|
2627 |
+
"loss": 0.029,
|
2628 |
+
"step": 3740
|
2629 |
+
},
|
2630 |
+
{
|
2631 |
+
"epoch": 7.653061224489796,
|
2632 |
+
"grad_norm": 0.3924682140350342,
|
2633 |
+
"learning_rate": 1.4341748249245623e-05,
|
2634 |
+
"loss": 0.0317,
|
2635 |
+
"step": 3750
|
2636 |
+
},
|
2637 |
+
{
|
2638 |
+
"epoch": 7.673469387755102,
|
2639 |
+
"grad_norm": 0.30662891268730164,
|
2640 |
+
"learning_rate": 1.410601598117246e-05,
|
2641 |
+
"loss": 0.0284,
|
2642 |
+
"step": 3760
|
2643 |
+
},
|
2644 |
+
{
|
2645 |
+
"epoch": 7.6938775510204085,
|
2646 |
+
"grad_norm": 0.22942852973937988,
|
2647 |
+
"learning_rate": 1.3871918572434345e-05,
|
2648 |
+
"loss": 0.0328,
|
2649 |
+
"step": 3770
|
2650 |
+
},
|
2651 |
+
{
|
2652 |
+
"epoch": 7.714285714285714,
|
2653 |
+
"grad_norm": 0.4252789616584778,
|
2654 |
+
"learning_rate": 1.3639466685440132e-05,
|
2655 |
+
"loss": 0.0322,
|
2656 |
+
"step": 3780
|
2657 |
+
},
|
2658 |
+
{
|
2659 |
+
"epoch": 7.73469387755102,
|
2660 |
+
"grad_norm": 0.40231239795684814,
|
2661 |
+
"learning_rate": 1.3408670907650322e-05,
|
2662 |
+
"loss": 0.0276,
|
2663 |
+
"step": 3790
|
2664 |
+
},
|
2665 |
+
{
|
2666 |
+
"epoch": 7.755102040816326,
|
2667 |
+
"grad_norm": 0.25347328186035156,
|
2668 |
+
"learning_rate": 1.3179541751094787e-05,
|
2669 |
+
"loss": 0.0277,
|
2670 |
+
"step": 3800
|
2671 |
+
},
|
2672 |
+
{
|
2673 |
+
"epoch": 7.775510204081632,
|
2674 |
+
"grad_norm": 0.3201022148132324,
|
2675 |
+
"learning_rate": 1.2952089651893945e-05,
|
2676 |
+
"loss": 0.0262,
|
2677 |
+
"step": 3810
|
2678 |
+
},
|
2679 |
+
{
|
2680 |
+
"epoch": 7.795918367346939,
|
2681 |
+
"grad_norm": 0.3606759011745453,
|
2682 |
+
"learning_rate": 1.2726324969783515e-05,
|
2683 |
+
"loss": 0.0304,
|
2684 |
+
"step": 3820
|
2685 |
+
},
|
2686 |
+
{
|
2687 |
+
"epoch": 7.816326530612245,
|
2688 |
+
"grad_norm": 0.3817797601222992,
|
2689 |
+
"learning_rate": 1.2502257987642591e-05,
|
2690 |
+
"loss": 0.0325,
|
2691 |
+
"step": 3830
|
2692 |
+
},
|
2693 |
+
{
|
2694 |
+
"epoch": 7.836734693877551,
|
2695 |
+
"grad_norm": 0.3720795512199402,
|
2696 |
+
"learning_rate": 1.2279898911025333e-05,
|
2697 |
+
"loss": 0.027,
|
2698 |
+
"step": 3840
|
2699 |
+
},
|
2700 |
+
{
|
2701 |
+
"epoch": 7.857142857142857,
|
2702 |
+
"grad_norm": 0.5197123885154724,
|
2703 |
+
"learning_rate": 1.2059257867696073e-05,
|
2704 |
+
"loss": 0.0284,
|
2705 |
+
"step": 3850
|
2706 |
+
},
|
2707 |
+
{
|
2708 |
+
"epoch": 7.877551020408164,
|
2709 |
+
"grad_norm": 0.26620540022850037,
|
2710 |
+
"learning_rate": 1.1840344907168122e-05,
|
2711 |
+
"loss": 0.0236,
|
2712 |
+
"step": 3860
|
2713 |
+
},
|
2714 |
+
{
|
2715 |
+
"epoch": 7.8979591836734695,
|
2716 |
+
"grad_norm": 0.5102793574333191,
|
2717 |
+
"learning_rate": 1.1623170000245975e-05,
|
2718 |
+
"loss": 0.033,
|
2719 |
+
"step": 3870
|
2720 |
+
},
|
2721 |
+
{
|
2722 |
+
"epoch": 7.918367346938775,
|
2723 |
+
"grad_norm": 0.47227033972740173,
|
2724 |
+
"learning_rate": 1.1407743038571172e-05,
|
2725 |
+
"loss": 0.0328,
|
2726 |
+
"step": 3880
|
2727 |
+
},
|
2728 |
+
{
|
2729 |
+
"epoch": 7.938775510204081,
|
2730 |
+
"grad_norm": 0.17635785043239594,
|
2731 |
+
"learning_rate": 1.1194073834171804e-05,
|
2732 |
+
"loss": 0.0223,
|
2733 |
+
"step": 3890
|
2734 |
+
},
|
2735 |
+
{
|
2736 |
+
"epoch": 7.959183673469388,
|
2737 |
+
"grad_norm": 0.6158728003501892,
|
2738 |
+
"learning_rate": 1.0982172119015593e-05,
|
2739 |
+
"loss": 0.0322,
|
2740 |
+
"step": 3900
|
2741 |
+
},
|
2742 |
+
{
|
2743 |
+
"epoch": 7.979591836734694,
|
2744 |
+
"grad_norm": 0.29061493277549744,
|
2745 |
+
"learning_rate": 1.0772047544566589e-05,
|
2746 |
+
"loss": 0.0269,
|
2747 |
+
"step": 3910
|
2748 |
+
},
|
2749 |
+
{
|
2750 |
+
"epoch": 8.0,
|
2751 |
+
"grad_norm": 0.36250999569892883,
|
2752 |
+
"learning_rate": 1.056370968134564e-05,
|
2753 |
+
"loss": 0.0258,
|
2754 |
+
"step": 3920
|
2755 |
+
},
|
2756 |
+
{
|
2757 |
+
"epoch": 8.020408163265307,
|
2758 |
+
"grad_norm": 0.288967490196228,
|
2759 |
+
"learning_rate": 1.0357168018494462e-05,
|
2760 |
+
"loss": 0.0242,
|
2761 |
+
"step": 3930
|
2762 |
+
},
|
2763 |
+
{
|
2764 |
+
"epoch": 8.040816326530612,
|
2765 |
+
"grad_norm": 0.35259634256362915,
|
2766 |
+
"learning_rate": 1.0152431963343384e-05,
|
2767 |
+
"loss": 0.0248,
|
2768 |
+
"step": 3940
|
2769 |
+
},
|
2770 |
+
{
|
2771 |
+
"epoch": 8.061224489795919,
|
2772 |
+
"grad_norm": 0.2822493314743042,
|
2773 |
+
"learning_rate": 9.949510840982985e-06,
|
2774 |
+
"loss": 0.0249,
|
2775 |
+
"step": 3950
|
2776 |
+
},
|
2777 |
+
{
|
2778 |
+
"epoch": 8.081632653061224,
|
2779 |
+
"grad_norm": 0.3533857762813568,
|
2780 |
+
"learning_rate": 9.748413893839231e-06,
|
2781 |
+
"loss": 0.0272,
|
2782 |
+
"step": 3960
|
2783 |
+
},
|
2784 |
+
{
|
2785 |
+
"epoch": 8.10204081632653,
|
2786 |
+
"grad_norm": 0.4431054890155792,
|
2787 |
+
"learning_rate": 9.549150281252633e-06,
|
2788 |
+
"loss": 0.0276,
|
2789 |
+
"step": 3970
|
2790 |
+
},
|
2791 |
+
{
|
2792 |
+
"epoch": 8.122448979591837,
|
2793 |
+
"grad_norm": 0.5681186318397522,
|
2794 |
+
"learning_rate": 9.351729079061005e-06,
|
2795 |
+
"loss": 0.0263,
|
2796 |
+
"step": 3980
|
2797 |
+
},
|
2798 |
+
{
|
2799 |
+
"epoch": 8.142857142857142,
|
2800 |
+
"grad_norm": 0.26654306054115295,
|
2801 |
+
"learning_rate": 9.156159279186078e-06,
|
2802 |
+
"loss": 0.0254,
|
2803 |
+
"step": 3990
|
2804 |
+
},
|
2805 |
+
{
|
2806 |
+
"epoch": 8.16326530612245,
|
2807 |
+
"grad_norm": 0.9877157211303711,
|
2808 |
+
"learning_rate": 8.96244978922397e-06,
|
2809 |
+
"loss": 0.0239,
|
2810 |
+
"step": 4000
|
2811 |
+
},
|
2812 |
+
{
|
2813 |
+
"epoch": 8.183673469387756,
|
2814 |
+
"grad_norm": 0.30778566002845764,
|
2815 |
+
"learning_rate": 8.770609432039505e-06,
|
2816 |
+
"loss": 0.0281,
|
2817 |
+
"step": 4010
|
2818 |
+
},
|
2819 |
+
{
|
2820 |
+
"epoch": 8.204081632653061,
|
2821 |
+
"grad_norm": 0.5464140176773071,
|
2822 |
+
"learning_rate": 8.580646945364256e-06,
|
2823 |
+
"loss": 0.0243,
|
2824 |
+
"step": 4020
|
2825 |
+
},
|
2826 |
+
{
|
2827 |
+
"epoch": 8.224489795918368,
|
2828 |
+
"grad_norm": 0.34788644313812256,
|
2829 |
+
"learning_rate": 8.392570981398695e-06,
|
2830 |
+
"loss": 0.0307,
|
2831 |
+
"step": 4030
|
2832 |
+
},
|
2833 |
+
{
|
2834 |
+
"epoch": 8.244897959183673,
|
2835 |
+
"grad_norm": 0.31038105487823486,
|
2836 |
+
"learning_rate": 8.206390106418027e-06,
|
2837 |
+
"loss": 0.0249,
|
2838 |
+
"step": 4040
|
2839 |
+
},
|
2840 |
+
{
|
2841 |
+
"epoch": 8.26530612244898,
|
2842 |
+
"grad_norm": 0.2261275053024292,
|
2843 |
+
"learning_rate": 8.022112800382059e-06,
|
2844 |
+
"loss": 0.0234,
|
2845 |
+
"step": 4050
|
2846 |
+
},
|
2847 |
+
{
|
2848 |
+
"epoch": 8.285714285714286,
|
2849 |
+
"grad_norm": 0.2555391192436218,
|
2850 |
+
"learning_rate": 7.839747456548918e-06,
|
2851 |
+
"loss": 0.0298,
|
2852 |
+
"step": 4060
|
2853 |
+
},
|
2854 |
+
{
|
2855 |
+
"epoch": 8.306122448979592,
|
2856 |
+
"grad_norm": 0.2868110239505768,
|
2857 |
+
"learning_rate": 7.659302381092842e-06,
|
2858 |
+
"loss": 0.0295,
|
2859 |
+
"step": 4070
|
2860 |
+
},
|
2861 |
+
{
|
2862 |
+
"epoch": 8.326530612244898,
|
2863 |
+
"grad_norm": 0.2840665578842163,
|
2864 |
+
"learning_rate": 7.480785792725809e-06,
|
2865 |
+
"loss": 0.0214,
|
2866 |
+
"step": 4080
|
2867 |
+
},
|
2868 |
+
{
|
2869 |
+
"epoch": 8.346938775510203,
|
2870 |
+
"grad_norm": 0.3482186198234558,
|
2871 |
+
"learning_rate": 7.304205822323173e-06,
|
2872 |
+
"loss": 0.0265,
|
2873 |
+
"step": 4090
|
2874 |
+
},
|
2875 |
+
{
|
2876 |
+
"epoch": 8.36734693877551,
|
2877 |
+
"grad_norm": 0.3534049093723297,
|
2878 |
+
"learning_rate": 7.129570512553407e-06,
|
2879 |
+
"loss": 0.0267,
|
2880 |
+
"step": 4100
|
2881 |
+
},
|
2882 |
+
{
|
2883 |
+
"epoch": 8.387755102040817,
|
2884 |
+
"grad_norm": 0.2970373034477234,
|
2885 |
+
"learning_rate": 6.95688781751172e-06,
|
2886 |
+
"loss": 0.0271,
|
2887 |
+
"step": 4110
|
2888 |
+
},
|
2889 |
+
{
|
2890 |
+
"epoch": 8.408163265306122,
|
2891 |
+
"grad_norm": 0.3268330693244934,
|
2892 |
+
"learning_rate": 6.786165602357808e-06,
|
2893 |
+
"loss": 0.0275,
|
2894 |
+
"step": 4120
|
2895 |
+
},
|
2896 |
+
{
|
2897 |
+
"epoch": 8.428571428571429,
|
2898 |
+
"grad_norm": 0.5232375264167786,
|
2899 |
+
"learning_rate": 6.617411642957566e-06,
|
2900 |
+
"loss": 0.0248,
|
2901 |
+
"step": 4130
|
2902 |
+
},
|
2903 |
+
{
|
2904 |
+
"epoch": 8.448979591836734,
|
2905 |
+
"grad_norm": 0.37294477224349976,
|
2906 |
+
"learning_rate": 6.450633625529001e-06,
|
2907 |
+
"loss": 0.0283,
|
2908 |
+
"step": 4140
|
2909 |
+
},
|
2910 |
+
{
|
2911 |
+
"epoch": 8.46938775510204,
|
2912 |
+
"grad_norm": 0.3570898771286011,
|
2913 |
+
"learning_rate": 6.285839146292111e-06,
|
2914 |
+
"loss": 0.0276,
|
2915 |
+
"step": 4150
|
2916 |
+
},
|
2917 |
+
{
|
2918 |
+
"epoch": 8.489795918367347,
|
2919 |
+
"grad_norm": 0.28744009137153625,
|
2920 |
+
"learning_rate": 6.12303571112286e-06,
|
2921 |
+
"loss": 0.0203,
|
2922 |
+
"step": 4160
|
2923 |
+
},
|
2924 |
+
{
|
2925 |
+
"epoch": 8.510204081632653,
|
2926 |
+
"grad_norm": 0.4438728988170624,
|
2927 |
+
"learning_rate": 5.962230735211383e-06,
|
2928 |
+
"loss": 0.026,
|
2929 |
+
"step": 4170
|
2930 |
+
},
|
2931 |
+
{
|
2932 |
+
"epoch": 8.53061224489796,
|
2933 |
+
"grad_norm": 0.24973392486572266,
|
2934 |
+
"learning_rate": 5.803431542724192e-06,
|
2935 |
+
"loss": 0.0284,
|
2936 |
+
"step": 4180
|
2937 |
+
},
|
2938 |
+
{
|
2939 |
+
"epoch": 8.551020408163264,
|
2940 |
+
"grad_norm": 0.5999603271484375,
|
2941 |
+
"learning_rate": 5.646645366470621e-06,
|
2942 |
+
"loss": 0.0285,
|
2943 |
+
"step": 4190
|
2944 |
+
},
|
2945 |
+
{
|
2946 |
+
"epoch": 8.571428571428571,
|
2947 |
+
"grad_norm": 0.3535723090171814,
|
2948 |
+
"learning_rate": 5.491879347573353e-06,
|
2949 |
+
"loss": 0.0261,
|
2950 |
+
"step": 4200
|
2951 |
+
},
|
2952 |
+
{
|
2953 |
+
"epoch": 8.591836734693878,
|
2954 |
+
"grad_norm": 0.22815968096256256,
|
2955 |
+
"learning_rate": 5.3391405351432e-06,
|
2956 |
+
"loss": 0.0223,
|
2957 |
+
"step": 4210
|
2958 |
+
},
|
2959 |
+
{
|
2960 |
+
"epoch": 8.612244897959183,
|
2961 |
+
"grad_norm": 0.2869364023208618,
|
2962 |
+
"learning_rate": 5.188435885958037e-06,
|
2963 |
+
"loss": 0.0258,
|
2964 |
+
"step": 4220
|
2965 |
+
},
|
2966 |
+
{
|
2967 |
+
"epoch": 8.63265306122449,
|
2968 |
+
"grad_norm": 0.25079235434532166,
|
2969 |
+
"learning_rate": 5.039772264145887e-06,
|
2970 |
+
"loss": 0.0198,
|
2971 |
+
"step": 4230
|
2972 |
+
},
|
2973 |
+
{
|
2974 |
+
"epoch": 8.653061224489797,
|
2975 |
+
"grad_norm": 0.3177327811717987,
|
2976 |
+
"learning_rate": 4.8931564408723656e-06,
|
2977 |
+
"loss": 0.023,
|
2978 |
+
"step": 4240
|
2979 |
+
},
|
2980 |
+
{
|
2981 |
+
"epoch": 8.673469387755102,
|
2982 |
+
"grad_norm": 0.3457529544830322,
|
2983 |
+
"learning_rate": 4.748595094032221e-06,
|
2984 |
+
"loss": 0.0274,
|
2985 |
+
"step": 4250
|
2986 |
+
},
|
2987 |
+
{
|
2988 |
+
"epoch": 8.693877551020408,
|
2989 |
+
"grad_norm": 0.4598150849342346,
|
2990 |
+
"learning_rate": 4.606094807945199e-06,
|
2991 |
+
"loss": 0.0333,
|
2992 |
+
"step": 4260
|
2993 |
+
},
|
2994 |
+
{
|
2995 |
+
"epoch": 8.714285714285714,
|
2996 |
+
"grad_norm": 0.2446356564760208,
|
2997 |
+
"learning_rate": 4.465662073056109e-06,
|
2998 |
+
"loss": 0.0238,
|
2999 |
+
"step": 4270
|
3000 |
+
},
|
3001 |
+
{
|
3002 |
+
"epoch": 8.73469387755102,
|
3003 |
+
"grad_norm": 0.16839979588985443,
|
3004 |
+
"learning_rate": 4.327303285639262e-06,
|
3005 |
+
"loss": 0.0218,
|
3006 |
+
"step": 4280
|
3007 |
+
},
|
3008 |
+
{
|
3009 |
+
"epoch": 8.755102040816327,
|
3010 |
+
"grad_norm": 0.2948671281337738,
|
3011 |
+
"learning_rate": 4.1910247475071104e-06,
|
3012 |
+
"loss": 0.0262,
|
3013 |
+
"step": 4290
|
3014 |
+
},
|
3015 |
+
{
|
3016 |
+
"epoch": 8.775510204081632,
|
3017 |
+
"grad_norm": 0.28787535429000854,
|
3018 |
+
"learning_rate": 4.0568326657231985e-06,
|
3019 |
+
"loss": 0.0255,
|
3020 |
+
"step": 4300
|
3021 |
+
},
|
3022 |
+
{
|
3023 |
+
"epoch": 8.795918367346939,
|
3024 |
+
"grad_norm": 0.25958192348480225,
|
3025 |
+
"learning_rate": 3.924733152319493e-06,
|
3026 |
+
"loss": 0.0207,
|
3027 |
+
"step": 4310
|
3028 |
+
},
|
3029 |
+
{
|
3030 |
+
"epoch": 8.816326530612244,
|
3031 |
+
"grad_norm": 0.3317425549030304,
|
3032 |
+
"learning_rate": 3.794732224017994e-06,
|
3033 |
+
"loss": 0.0268,
|
3034 |
+
"step": 4320
|
3035 |
+
},
|
3036 |
+
{
|
3037 |
+
"epoch": 8.83673469387755,
|
3038 |
+
"grad_norm": 0.3428516685962677,
|
3039 |
+
"learning_rate": 3.6668358019566305e-06,
|
3040 |
+
"loss": 0.0212,
|
3041 |
+
"step": 4330
|
3042 |
+
},
|
3043 |
+
{
|
3044 |
+
"epoch": 8.857142857142858,
|
3045 |
+
"grad_norm": 0.365178644657135,
|
3046 |
+
"learning_rate": 3.541049711419664e-06,
|
3047 |
+
"loss": 0.0219,
|
3048 |
+
"step": 4340
|
3049 |
+
},
|
3050 |
+
{
|
3051 |
+
"epoch": 8.877551020408163,
|
3052 |
+
"grad_norm": 0.2565886378288269,
|
3053 |
+
"learning_rate": 3.417379681572297e-06,
|
3054 |
+
"loss": 0.0244,
|
3055 |
+
"step": 4350
|
3056 |
+
},
|
3057 |
+
{
|
3058 |
+
"epoch": 8.89795918367347,
|
3059 |
+
"grad_norm": 0.3833511471748352,
|
3060 |
+
"learning_rate": 3.295831345199746e-06,
|
3061 |
+
"loss": 0.021,
|
3062 |
+
"step": 4360
|
3063 |
+
},
|
3064 |
+
{
|
3065 |
+
"epoch": 8.918367346938776,
|
3066 |
+
"grad_norm": 0.2949051260948181,
|
3067 |
+
"learning_rate": 3.1764102384507e-06,
|
3068 |
+
"loss": 0.0259,
|
3069 |
+
"step": 4370
|
3070 |
+
},
|
3071 |
+
{
|
3072 |
+
"epoch": 8.938775510204081,
|
3073 |
+
"grad_norm": 0.5416157841682434,
|
3074 |
+
"learning_rate": 3.059121800585152e-06,
|
3075 |
+
"loss": 0.0196,
|
3076 |
+
"step": 4380
|
3077 |
+
},
|
3078 |
+
{
|
3079 |
+
"epoch": 8.959183673469388,
|
3080 |
+
"grad_norm": 0.327238529920578,
|
3081 |
+
"learning_rate": 2.9439713737266504e-06,
|
3082 |
+
"loss": 0.0256,
|
3083 |
+
"step": 4390
|
3084 |
+
},
|
3085 |
+
{
|
3086 |
+
"epoch": 8.979591836734693,
|
3087 |
+
"grad_norm": 0.4477047622203827,
|
3088 |
+
"learning_rate": 2.8309642026190075e-06,
|
3089 |
+
"loss": 0.0275,
|
3090 |
+
"step": 4400
|
3091 |
+
},
|
3092 |
+
{
|
3093 |
+
"epoch": 9.0,
|
3094 |
+
"grad_norm": 0.42429378628730774,
|
3095 |
+
"learning_rate": 2.720105434387382e-06,
|
3096 |
+
"loss": 0.0292,
|
3097 |
+
"step": 4410
|
3098 |
+
},
|
3099 |
+
{
|
3100 |
+
"epoch": 9.020408163265307,
|
3101 |
+
"grad_norm": 0.35936760902404785,
|
3102 |
+
"learning_rate": 2.6114001183038806e-06,
|
3103 |
+
"loss": 0.0228,
|
3104 |
+
"step": 4420
|
3105 |
+
},
|
3106 |
+
{
|
3107 |
+
"epoch": 9.040816326530612,
|
3108 |
+
"grad_norm": 0.4515119194984436,
|
3109 |
+
"learning_rate": 2.5048532055575615e-06,
|
3110 |
+
"loss": 0.0275,
|
3111 |
+
"step": 4430
|
3112 |
+
},
|
3113 |
+
{
|
3114 |
+
"epoch": 9.061224489795919,
|
3115 |
+
"grad_norm": 0.28619876503944397,
|
3116 |
+
"learning_rate": 2.400469549028922e-06,
|
3117 |
+
"loss": 0.0256,
|
3118 |
+
"step": 4440
|
3119 |
+
},
|
3120 |
+
{
|
3121 |
+
"epoch": 9.081632653061224,
|
3122 |
+
"grad_norm": 0.3474380075931549,
|
3123 |
+
"learning_rate": 2.2982539030688475e-06,
|
3124 |
+
"loss": 0.0307,
|
3125 |
+
"step": 4450
|
3126 |
+
},
|
3127 |
+
{
|
3128 |
+
"epoch": 9.10204081632653,
|
3129 |
+
"grad_norm": 0.21155385673046112,
|
3130 |
+
"learning_rate": 2.1982109232821178e-06,
|
3131 |
+
"loss": 0.0246,
|
3132 |
+
"step": 4460
|
3133 |
+
},
|
3134 |
+
{
|
3135 |
+
"epoch": 9.122448979591837,
|
3136 |
+
"grad_norm": 0.33730846643447876,
|
3137 |
+
"learning_rate": 2.1003451663153095e-06,
|
3138 |
+
"loss": 0.0203,
|
3139 |
+
"step": 4470
|
3140 |
+
},
|
3141 |
+
{
|
3142 |
+
"epoch": 9.142857142857142,
|
3143 |
+
"grad_norm": 0.3035857379436493,
|
3144 |
+
"learning_rate": 2.0046610896492624e-06,
|
3145 |
+
"loss": 0.021,
|
3146 |
+
"step": 4480
|
3147 |
+
},
|
3148 |
+
{
|
3149 |
+
"epoch": 9.16326530612245,
|
3150 |
+
"grad_norm": 0.7528371214866638,
|
3151 |
+
"learning_rate": 1.9111630513960755e-06,
|
3152 |
+
"loss": 0.0211,
|
3153 |
+
"step": 4490
|
3154 |
+
},
|
3155 |
+
{
|
3156 |
+
"epoch": 9.183673469387756,
|
3157 |
+
"grad_norm": 0.3337666392326355,
|
3158 |
+
"learning_rate": 1.8198553101005987e-06,
|
3159 |
+
"loss": 0.0278,
|
3160 |
+
"step": 4500
|
3161 |
+
},
|
3162 |
+
{
|
3163 |
+
"epoch": 9.204081632653061,
|
3164 |
+
"grad_norm": 0.7967633605003357,
|
3165 |
+
"learning_rate": 1.7307420245464667e-06,
|
3166 |
+
"loss": 0.0315,
|
3167 |
+
"step": 4510
|
3168 |
+
},
|
3169 |
+
{
|
3170 |
+
"epoch": 9.224489795918368,
|
3171 |
+
"grad_norm": 0.22648100554943085,
|
3172 |
+
"learning_rate": 1.643827253566671e-06,
|
3173 |
+
"loss": 0.0268,
|
3174 |
+
"step": 4520
|
3175 |
+
},
|
3176 |
+
{
|
3177 |
+
"epoch": 9.244897959183673,
|
3178 |
+
"grad_norm": 0.2347017526626587,
|
3179 |
+
"learning_rate": 1.5591149558587038e-06,
|
3180 |
+
"loss": 0.0232,
|
3181 |
+
"step": 4530
|
3182 |
+
},
|
3183 |
+
{
|
3184 |
+
"epoch": 9.26530612244898,
|
3185 |
+
"grad_norm": 0.38450396060943604,
|
3186 |
+
"learning_rate": 1.4766089898042678e-06,
|
3187 |
+
"loss": 0.023,
|
3188 |
+
"step": 4540
|
3189 |
+
},
|
3190 |
+
{
|
3191 |
+
"epoch": 9.285714285714286,
|
3192 |
+
"grad_norm": 0.49666064977645874,
|
3193 |
+
"learning_rate": 1.3963131132935015e-06,
|
3194 |
+
"loss": 0.021,
|
3195 |
+
"step": 4550
|
3196 |
+
},
|
3197 |
+
{
|
3198 |
+
"epoch": 9.306122448979592,
|
3199 |
+
"grad_norm": 0.35134318470954895,
|
3200 |
+
"learning_rate": 1.3182309835538365e-06,
|
3201 |
+
"loss": 0.0213,
|
3202 |
+
"step": 4560
|
3203 |
+
},
|
3204 |
+
{
|
3205 |
+
"epoch": 9.326530612244898,
|
3206 |
+
"grad_norm": 0.24812936782836914,
|
3207 |
+
"learning_rate": 1.2423661569834489e-06,
|
3208 |
+
"loss": 0.0248,
|
3209 |
+
"step": 4570
|
3210 |
+
},
|
3211 |
+
{
|
3212 |
+
"epoch": 9.346938775510203,
|
3213 |
+
"grad_norm": 0.2637794315814972,
|
3214 |
+
"learning_rate": 1.168722088989227e-06,
|
3215 |
+
"loss": 0.022,
|
3216 |
+
"step": 4580
|
3217 |
+
},
|
3218 |
+
{
|
3219 |
+
"epoch": 9.36734693877551,
|
3220 |
+
"grad_norm": 0.25408467650413513,
|
3221 |
+
"learning_rate": 1.0973021338294143e-06,
|
3222 |
+
"loss": 0.025,
|
3223 |
+
"step": 4590
|
3224 |
+
},
|
3225 |
+
{
|
3226 |
+
"epoch": 9.387755102040817,
|
3227 |
+
"grad_norm": 0.2591371536254883,
|
3228 |
+
"learning_rate": 1.0281095444608423e-06,
|
3229 |
+
"loss": 0.0228,
|
3230 |
+
"step": 4600
|
3231 |
+
},
|
3232 |
+
{
|
3233 |
+
"epoch": 9.408163265306122,
|
3234 |
+
"grad_norm": 0.2628161609172821,
|
3235 |
+
"learning_rate": 9.611474723907666e-07,
|
3236 |
+
"loss": 0.0274,
|
3237 |
+
"step": 4610
|
3238 |
+
},
|
3239 |
+
{
|
3240 |
+
"epoch": 9.428571428571429,
|
3241 |
+
"grad_norm": 0.24642103910446167,
|
3242 |
+
"learning_rate": 8.964189675332912e-07,
|
3243 |
+
"loss": 0.0218,
|
3244 |
+
"step": 4620
|
3245 |
+
},
|
3246 |
+
{
|
3247 |
+
"epoch": 9.448979591836734,
|
3248 |
+
"grad_norm": 0.2833108901977539,
|
3249 |
+
"learning_rate": 8.33926978070504e-07,
|
3250 |
+
"loss": 0.0242,
|
3251 |
+
"step": 4630
|
3252 |
+
},
|
3253 |
+
{
|
3254 |
+
"epoch": 9.46938775510204,
|
3255 |
+
"grad_norm": 0.17190630733966827,
|
3256 |
+
"learning_rate": 7.736743503181543e-07,
|
3257 |
+
"loss": 0.0226,
|
3258 |
+
"step": 4640
|
3259 |
+
},
|
3260 |
+
{
|
3261 |
+
"epoch": 9.489795918367347,
|
3262 |
+
"grad_norm": 0.28679272532463074,
|
3263 |
+
"learning_rate": 7.156638285960471e-07,
|
3264 |
+
"loss": 0.0245,
|
3265 |
+
"step": 4650
|
3266 |
+
},
|
3267 |
+
{
|
3268 |
+
"epoch": 9.510204081632653,
|
3269 |
+
"grad_norm": 0.3385928273200989,
|
3270 |
+
"learning_rate": 6.598980551030143e-07,
|
3271 |
+
"loss": 0.0271,
|
3272 |
+
"step": 4660
|
3273 |
+
},
|
3274 |
+
{
|
3275 |
+
"epoch": 9.53061224489796,
|
3276 |
+
"grad_norm": 0.3173186779022217,
|
3277 |
+
"learning_rate": 6.063795697966057e-07,
|
3278 |
+
"loss": 0.0251,
|
3279 |
+
"step": 4670
|
3280 |
+
},
|
3281 |
+
{
|
3282 |
+
"epoch": 9.551020408163264,
|
3283 |
+
"grad_norm": 0.28356119990348816,
|
3284 |
+
"learning_rate": 5.551108102773761e-07,
|
3285 |
+
"loss": 0.0207,
|
3286 |
+
"step": 4680
|
3287 |
+
},
|
3288 |
+
{
|
3289 |
+
"epoch": 9.571428571428571,
|
3290 |
+
"grad_norm": 0.3257679045200348,
|
3291 |
+
"learning_rate": 5.060941116778572e-07,
|
3292 |
+
"loss": 0.0272,
|
3293 |
+
"step": 4690
|
3294 |
+
},
|
3295 |
+
{
|
3296 |
+
"epoch": 9.591836734693878,
|
3297 |
+
"grad_norm": 0.36799612641334534,
|
3298 |
+
"learning_rate": 4.5933170655622083e-07,
|
3299 |
+
"loss": 0.0263,
|
3300 |
+
"step": 4700
|
3301 |
+
},
|
3302 |
+
{
|
3303 |
+
"epoch": 9.612244897959183,
|
3304 |
+
"grad_norm": 0.5192159414291382,
|
3305 |
+
"learning_rate": 4.1482572479458193e-07,
|
3306 |
+
"loss": 0.0241,
|
3307 |
+
"step": 4710
|
3308 |
+
},
|
3309 |
+
{
|
3310 |
+
"epoch": 9.63265306122449,
|
3311 |
+
"grad_norm": 0.24428972601890564,
|
3312 |
+
"learning_rate": 3.725781935019934e-07,
|
3313 |
+
"loss": 0.025,
|
3314 |
+
"step": 4720
|
3315 |
+
},
|
3316 |
+
{
|
3317 |
+
"epoch": 9.653061224489797,
|
3318 |
+
"grad_norm": 0.35796719789505005,
|
3319 |
+
"learning_rate": 3.3259103692209747e-07,
|
3320 |
+
"loss": 0.0223,
|
3321 |
+
"step": 4730
|
3322 |
+
},
|
3323 |
+
{
|
3324 |
+
"epoch": 9.673469387755102,
|
3325 |
+
"grad_norm": 0.264077365398407,
|
3326 |
+
"learning_rate": 2.9486607634551797e-07,
|
3327 |
+
"loss": 0.0225,
|
3328 |
+
"step": 4740
|
3329 |
+
},
|
3330 |
+
{
|
3331 |
+
"epoch": 9.693877551020408,
|
3332 |
+
"grad_norm": 0.6665878295898438,
|
3333 |
+
"learning_rate": 2.594050300268769e-07,
|
3334 |
+
"loss": 0.0228,
|
3335 |
+
"step": 4750
|
3336 |
+
},
|
3337 |
+
{
|
3338 |
+
"epoch": 9.714285714285714,
|
3339 |
+
"grad_norm": 0.3441040813922882,
|
3340 |
+
"learning_rate": 2.2620951310653448e-07,
|
3341 |
+
"loss": 0.0259,
|
3342 |
+
"step": 4760
|
3343 |
+
},
|
3344 |
+
{
|
3345 |
+
"epoch": 9.73469387755102,
|
3346 |
+
"grad_norm": 0.322299599647522,
|
3347 |
+
"learning_rate": 1.9528103753704842e-07,
|
3348 |
+
"loss": 0.0254,
|
3349 |
+
"step": 4770
|
3350 |
+
},
|
3351 |
+
{
|
3352 |
+
"epoch": 9.755102040816327,
|
3353 |
+
"grad_norm": 0.3630422055721283,
|
3354 |
+
"learning_rate": 1.666210120142897e-07,
|
3355 |
+
"loss": 0.0256,
|
3356 |
+
"step": 4780
|
3357 |
+
},
|
3358 |
+
{
|
3359 |
+
"epoch": 9.775510204081632,
|
3360 |
+
"grad_norm": 0.298095166683197,
|
3361 |
+
"learning_rate": 1.4023074191327756e-07,
|
3362 |
+
"loss": 0.0231,
|
3363 |
+
"step": 4790
|
3364 |
+
},
|
3365 |
+
{
|
3366 |
+
"epoch": 9.795918367346939,
|
3367 |
+
"grad_norm": 0.30586209893226624,
|
3368 |
+
"learning_rate": 1.1611142922874352e-07,
|
3369 |
+
"loss": 0.0199,
|
3370 |
+
"step": 4800
|
3371 |
+
},
|
3372 |
+
{
|
3373 |
+
"epoch": 9.816326530612244,
|
3374 |
+
"grad_norm": 0.4290463924407959,
|
3375 |
+
"learning_rate": 9.426417252035858e-08,
|
3376 |
+
"loss": 0.0234,
|
3377 |
+
"step": 4810
|
3378 |
+
},
|
3379 |
+
{
|
3380 |
+
"epoch": 9.83673469387755,
|
3381 |
+
"grad_norm": 0.2478075474500656,
|
3382 |
+
"learning_rate": 7.468996686271768e-08,
|
3383 |
+
"loss": 0.0234,
|
3384 |
+
"step": 4820
|
3385 |
+
},
|
3386 |
+
{
|
3387 |
+
"epoch": 9.857142857142858,
|
3388 |
+
"grad_norm": 0.40217646956443787,
|
3389 |
+
"learning_rate": 5.73897038000093e-08,
|
3390 |
+
"loss": 0.0252,
|
3391 |
+
"step": 4830
|
3392 |
+
},
|
3393 |
+
{
|
3394 |
+
"epoch": 9.877551020408163,
|
3395 |
+
"grad_norm": 0.3873520791530609,
|
3396 |
+
"learning_rate": 4.236417130540904e-08,
|
3397 |
+
"loss": 0.0325,
|
3398 |
+
"step": 4840
|
3399 |
+
},
|
3400 |
+
{
|
3401 |
+
"epoch": 9.89795918367347,
|
3402 |
+
"grad_norm": 0.32106974720954895,
|
3403 |
+
"learning_rate": 2.961405374519166e-08,
|
3404 |
+
"loss": 0.0237,
|
3405 |
+
"step": 4850
|
3406 |
+
},
|
3407 |
+
{
|
3408 |
+
"epoch": 9.918367346938776,
|
3409 |
+
"grad_norm": 0.28574123978614807,
|
3410 |
+
"learning_rate": 1.9139931847556026e-08,
|
3411 |
+
"loss": 0.0239,
|
3412 |
+
"step": 4860
|
3413 |
+
},
|
3414 |
+
{
|
3415 |
+
"epoch": 9.938775510204081,
|
3416 |
+
"grad_norm": 0.19580473005771637,
|
3417 |
+
"learning_rate": 1.0942282676174032e-08,
|
3418 |
+
"loss": 0.026,
|
3419 |
+
"step": 4870
|
3420 |
+
},
|
3421 |
+
{
|
3422 |
+
"epoch": 9.959183673469388,
|
3423 |
+
"grad_norm": 0.2628851532936096,
|
3424 |
+
"learning_rate": 5.021479608474655e-09,
|
3425 |
+
"loss": 0.0271,
|
3426 |
+
"step": 4880
|
3427 |
+
},
|
3428 |
+
{
|
3429 |
+
"epoch": 9.979591836734693,
|
3430 |
+
"grad_norm": 0.2211391031742096,
|
3431 |
+
"learning_rate": 1.377792318624227e-09,
|
3432 |
+
"loss": 0.0261,
|
3433 |
+
"step": 4890
|
3434 |
+
},
|
3435 |
+
{
|
3436 |
+
"epoch": 10.0,
|
3437 |
+
"grad_norm": 0.34537559747695923,
|
3438 |
+
"learning_rate": 1.1386765241816477e-11,
|
3439 |
+
"loss": 0.0276,
|
3440 |
+
"step": 4900
|
3441 |
+
},
|
3442 |
+
{
|
3443 |
+
"epoch": 10.0,
|
3444 |
+
"step": 4900,
|
3445 |
+
"total_flos": 0.0,
|
3446 |
+
"train_loss": 0.05459458509270026,
|
3447 |
+
"train_runtime": 4855.4962,
|
3448 |
+
"train_samples_per_second": 32.258,
|
3449 |
+
"train_steps_per_second": 1.009
|
3450 |
+
}
|
3451 |
+
],
|
3452 |
+
"logging_steps": 10,
|
3453 |
+
"max_steps": 4900,
|
3454 |
+
"num_input_tokens_seen": 0,
|
3455 |
+
"num_train_epochs": 10,
|
3456 |
+
"save_steps": 20000,
|
3457 |
+
"stateful_callbacks": {
|
3458 |
+
"TrainerControl": {
|
3459 |
+
"args": {
|
3460 |
+
"should_epoch_stop": false,
|
3461 |
+
"should_evaluate": false,
|
3462 |
+
"should_log": false,
|
3463 |
+
"should_save": true,
|
3464 |
+
"should_training_stop": true
|
3465 |
+
},
|
3466 |
+
"attributes": {}
|
3467 |
+
}
|
3468 |
+
},
|
3469 |
+
"total_flos": 0.0,
|
3470 |
+
"train_batch_size": 32,
|
3471 |
+
"trial_name": null,
|
3472 |
+
"trial_params": null
|
3473 |
+
}
|