Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1960 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1960 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 15.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2745,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0546448087431694,
|
| 13 |
+
"grad_norm": 14.413156509399414,
|
| 14 |
+
"learning_rate": 7.246376811594203e-06,
|
| 15 |
+
"loss": 1.7399,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.1092896174863388,
|
| 20 |
+
"grad_norm": 7.02054500579834,
|
| 21 |
+
"learning_rate": 1.4492753623188407e-05,
|
| 22 |
+
"loss": 1.2263,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.16393442622950818,
|
| 27 |
+
"grad_norm": 2.4340662956237793,
|
| 28 |
+
"learning_rate": 2.173913043478261e-05,
|
| 29 |
+
"loss": 0.401,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.2185792349726776,
|
| 34 |
+
"grad_norm": 1.5783411264419556,
|
| 35 |
+
"learning_rate": 2.8985507246376814e-05,
|
| 36 |
+
"loss": 0.2995,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.273224043715847,
|
| 41 |
+
"grad_norm": 1.2143406867980957,
|
| 42 |
+
"learning_rate": 3.6231884057971014e-05,
|
| 43 |
+
"loss": 0.224,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.32786885245901637,
|
| 48 |
+
"grad_norm": 1.4175114631652832,
|
| 49 |
+
"learning_rate": 4.347826086956522e-05,
|
| 50 |
+
"loss": 0.1867,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.3825136612021858,
|
| 55 |
+
"grad_norm": 1.3743549585342407,
|
| 56 |
+
"learning_rate": 5.072463768115943e-05,
|
| 57 |
+
"loss": 0.1733,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.4371584699453552,
|
| 62 |
+
"grad_norm": 1.4106934070587158,
|
| 63 |
+
"learning_rate": 5.797101449275363e-05,
|
| 64 |
+
"loss": 0.1607,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.4918032786885246,
|
| 69 |
+
"grad_norm": 0.7274103164672852,
|
| 70 |
+
"learning_rate": 6.521739130434783e-05,
|
| 71 |
+
"loss": 0.1481,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.546448087431694,
|
| 76 |
+
"grad_norm": 3.3031466007232666,
|
| 77 |
+
"learning_rate": 7.246376811594203e-05,
|
| 78 |
+
"loss": 0.1398,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.6010928961748634,
|
| 83 |
+
"grad_norm": 0.6688568592071533,
|
| 84 |
+
"learning_rate": 7.971014492753623e-05,
|
| 85 |
+
"loss": 0.131,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.6557377049180327,
|
| 90 |
+
"grad_norm": 0.6042072176933289,
|
| 91 |
+
"learning_rate": 8.695652173913044e-05,
|
| 92 |
+
"loss": 0.1303,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.7103825136612022,
|
| 97 |
+
"grad_norm": 0.9759267568588257,
|
| 98 |
+
"learning_rate": 9.420289855072463e-05,
|
| 99 |
+
"loss": 0.1184,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.7650273224043715,
|
| 104 |
+
"grad_norm": 1.2210289239883423,
|
| 105 |
+
"learning_rate": 9.999985478299589e-05,
|
| 106 |
+
"loss": 0.1227,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.819672131147541,
|
| 111 |
+
"grad_norm": 1.1991188526153564,
|
| 112 |
+
"learning_rate": 9.999477227642103e-05,
|
| 113 |
+
"loss": 0.1116,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.8743169398907104,
|
| 118 |
+
"grad_norm": 1.208275556564331,
|
| 119 |
+
"learning_rate": 9.998242976313776e-05,
|
| 120 |
+
"loss": 0.1081,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.9289617486338798,
|
| 125 |
+
"grad_norm": 0.8473228812217712,
|
| 126 |
+
"learning_rate": 9.996282903546813e-05,
|
| 127 |
+
"loss": 0.1,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.9836065573770492,
|
| 132 |
+
"grad_norm": 0.7348740100860596,
|
| 133 |
+
"learning_rate": 9.993597293973796e-05,
|
| 134 |
+
"loss": 0.1039,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 1.0382513661202186,
|
| 139 |
+
"grad_norm": 0.7729175090789795,
|
| 140 |
+
"learning_rate": 9.990186537586375e-05,
|
| 141 |
+
"loss": 0.1001,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 1.092896174863388,
|
| 146 |
+
"grad_norm": 1.019771695137024,
|
| 147 |
+
"learning_rate": 9.986051129678617e-05,
|
| 148 |
+
"loss": 0.0975,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 1.1475409836065573,
|
| 153 |
+
"grad_norm": 0.6440282464027405,
|
| 154 |
+
"learning_rate": 9.981191670775097e-05,
|
| 155 |
+
"loss": 0.0946,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 1.2021857923497268,
|
| 160 |
+
"grad_norm": 0.5440956950187683,
|
| 161 |
+
"learning_rate": 9.975608866543677e-05,
|
| 162 |
+
"loss": 0.0892,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 1.2568306010928962,
|
| 167 |
+
"grad_norm": 0.9462321996688843,
|
| 168 |
+
"learning_rate": 9.969303527693043e-05,
|
| 169 |
+
"loss": 0.09,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.3114754098360657,
|
| 174 |
+
"grad_norm": 0.7312270402908325,
|
| 175 |
+
"learning_rate": 9.962276569854977e-05,
|
| 176 |
+
"loss": 0.0841,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.366120218579235,
|
| 181 |
+
"grad_norm": 0.6993730664253235,
|
| 182 |
+
"learning_rate": 9.954529013451393e-05,
|
| 183 |
+
"loss": 0.0823,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.4207650273224044,
|
| 188 |
+
"grad_norm": 0.6108803749084473,
|
| 189 |
+
"learning_rate": 9.946061983546144e-05,
|
| 190 |
+
"loss": 0.0858,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 1.4754098360655736,
|
| 195 |
+
"grad_norm": 0.645415723323822,
|
| 196 |
+
"learning_rate": 9.936876709681668e-05,
|
| 197 |
+
"loss": 0.0883,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.530054644808743,
|
| 202 |
+
"grad_norm": 0.5133969783782959,
|
| 203 |
+
"learning_rate": 9.926974525700418e-05,
|
| 204 |
+
"loss": 0.0817,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 1.5846994535519126,
|
| 209 |
+
"grad_norm": 0.39707934856414795,
|
| 210 |
+
"learning_rate": 9.916356869551182e-05,
|
| 211 |
+
"loss": 0.0762,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.639344262295082,
|
| 216 |
+
"grad_norm": 0.43346527218818665,
|
| 217 |
+
"learning_rate": 9.905025283080265e-05,
|
| 218 |
+
"loss": 0.0826,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.6939890710382515,
|
| 223 |
+
"grad_norm": 0.8236570954322815,
|
| 224 |
+
"learning_rate": 9.892981411807587e-05,
|
| 225 |
+
"loss": 0.0782,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.748633879781421,
|
| 230 |
+
"grad_norm": 0.8405441045761108,
|
| 231 |
+
"learning_rate": 9.880227004687736e-05,
|
| 232 |
+
"loss": 0.0785,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.8032786885245902,
|
| 237 |
+
"grad_norm": 0.5183781385421753,
|
| 238 |
+
"learning_rate": 9.866763913855988e-05,
|
| 239 |
+
"loss": 0.0739,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.8579234972677594,
|
| 244 |
+
"grad_norm": 0.5411970615386963,
|
| 245 |
+
"learning_rate": 9.852594094359342e-05,
|
| 246 |
+
"loss": 0.0702,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.9125683060109289,
|
| 251 |
+
"grad_norm": 0.8432384133338928,
|
| 252 |
+
"learning_rate": 9.837719603872633e-05,
|
| 253 |
+
"loss": 0.084,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.9672131147540983,
|
| 258 |
+
"grad_norm": 0.8698591589927673,
|
| 259 |
+
"learning_rate": 9.82214260239971e-05,
|
| 260 |
+
"loss": 0.0778,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 2.021857923497268,
|
| 265 |
+
"grad_norm": 0.5935088396072388,
|
| 266 |
+
"learning_rate": 9.805865351959784e-05,
|
| 267 |
+
"loss": 0.0706,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 2.0765027322404372,
|
| 272 |
+
"grad_norm": 0.7685546278953552,
|
| 273 |
+
"learning_rate": 9.788890216258939e-05,
|
| 274 |
+
"loss": 0.0698,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 2.1311475409836067,
|
| 279 |
+
"grad_norm": 0.7633505463600159,
|
| 280 |
+
"learning_rate": 9.771219660346886e-05,
|
| 281 |
+
"loss": 0.0699,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 2.185792349726776,
|
| 286 |
+
"grad_norm": 0.31018251180648804,
|
| 287 |
+
"learning_rate": 9.752856250259006e-05,
|
| 288 |
+
"loss": 0.0804,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 2.240437158469945,
|
| 293 |
+
"grad_norm": 0.3807033598423004,
|
| 294 |
+
"learning_rate": 9.733802652643717e-05,
|
| 295 |
+
"loss": 0.0645,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 2.2950819672131146,
|
| 300 |
+
"grad_norm": 0.45358163118362427,
|
| 301 |
+
"learning_rate": 9.714061634375238e-05,
|
| 302 |
+
"loss": 0.066,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 2.349726775956284,
|
| 307 |
+
"grad_norm": 0.7382698655128479,
|
| 308 |
+
"learning_rate": 9.693636062151798e-05,
|
| 309 |
+
"loss": 0.0649,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 2.4043715846994536,
|
| 314 |
+
"grad_norm": 1.3578612804412842,
|
| 315 |
+
"learning_rate": 9.67252890207934e-05,
|
| 316 |
+
"loss": 0.0746,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 2.459016393442623,
|
| 321 |
+
"grad_norm": 1.0824800729751587,
|
| 322 |
+
"learning_rate": 9.650743219240813e-05,
|
| 323 |
+
"loss": 0.0695,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 2.5136612021857925,
|
| 328 |
+
"grad_norm": 0.6622858643531799,
|
| 329 |
+
"learning_rate": 9.628282177251054e-05,
|
| 330 |
+
"loss": 0.0717,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 2.5683060109289615,
|
| 335 |
+
"grad_norm": 0.5879955291748047,
|
| 336 |
+
"learning_rate": 9.605149037797396e-05,
|
| 337 |
+
"loss": 0.0659,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.6229508196721314,
|
| 342 |
+
"grad_norm": 0.6211952567100525,
|
| 343 |
+
"learning_rate": 9.581347160166023e-05,
|
| 344 |
+
"loss": 0.0651,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.6775956284153004,
|
| 349 |
+
"grad_norm": 0.5267804265022278,
|
| 350 |
+
"learning_rate": 9.55688000075414e-05,
|
| 351 |
+
"loss": 0.0628,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 2.73224043715847,
|
| 356 |
+
"grad_norm": 0.7821051478385925,
|
| 357 |
+
"learning_rate": 9.531751112568061e-05,
|
| 358 |
+
"loss": 0.0603,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 2.7868852459016393,
|
| 363 |
+
"grad_norm": 0.5416728258132935,
|
| 364 |
+
"learning_rate": 9.50596414470725e-05,
|
| 365 |
+
"loss": 0.0615,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 2.841530054644809,
|
| 370 |
+
"grad_norm": 0.7619845867156982,
|
| 371 |
+
"learning_rate": 9.479522841834423e-05,
|
| 372 |
+
"loss": 0.0653,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 2.8961748633879782,
|
| 377 |
+
"grad_norm": 1.21845543384552,
|
| 378 |
+
"learning_rate": 9.452431043631758e-05,
|
| 379 |
+
"loss": 0.0621,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 2.9508196721311473,
|
| 384 |
+
"grad_norm": 0.6109068989753723,
|
| 385 |
+
"learning_rate": 9.424692684243325e-05,
|
| 386 |
+
"loss": 0.0633,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 3.0054644808743167,
|
| 391 |
+
"grad_norm": 0.3331596851348877,
|
| 392 |
+
"learning_rate": 9.396311791703774e-05,
|
| 393 |
+
"loss": 0.057,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 3.060109289617486,
|
| 398 |
+
"grad_norm": 0.5244349837303162,
|
| 399 |
+
"learning_rate": 9.367292487353417e-05,
|
| 400 |
+
"loss": 0.064,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 3.1147540983606556,
|
| 405 |
+
"grad_norm": 0.771816611289978,
|
| 406 |
+
"learning_rate": 9.337638985239736e-05,
|
| 407 |
+
"loss": 0.0632,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 3.169398907103825,
|
| 412 |
+
"grad_norm": 0.49687984585762024,
|
| 413 |
+
"learning_rate": 9.307355591505442e-05,
|
| 414 |
+
"loss": 0.0599,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 3.2240437158469946,
|
| 419 |
+
"grad_norm": 0.41998782753944397,
|
| 420 |
+
"learning_rate": 9.276446703763159e-05,
|
| 421 |
+
"loss": 0.0616,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 3.278688524590164,
|
| 426 |
+
"grad_norm": 0.35080578923225403,
|
| 427 |
+
"learning_rate": 9.244916810456821e-05,
|
| 428 |
+
"loss": 0.0538,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 3.3333333333333335,
|
| 433 |
+
"grad_norm": 0.4821111559867859,
|
| 434 |
+
"learning_rate": 9.21277049020988e-05,
|
| 435 |
+
"loss": 0.0563,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 3.387978142076503,
|
| 440 |
+
"grad_norm": 0.9744987487792969,
|
| 441 |
+
"learning_rate": 9.180012411160426e-05,
|
| 442 |
+
"loss": 0.0591,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 3.442622950819672,
|
| 447 |
+
"grad_norm": 0.46575257182121277,
|
| 448 |
+
"learning_rate": 9.146647330283296e-05,
|
| 449 |
+
"loss": 0.0574,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 3.4972677595628414,
|
| 454 |
+
"grad_norm": 0.6140480041503906,
|
| 455 |
+
"learning_rate": 9.112680092699287e-05,
|
| 456 |
+
"loss": 0.0599,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 3.551912568306011,
|
| 461 |
+
"grad_norm": 0.5836021900177002,
|
| 462 |
+
"learning_rate": 9.078115630971574e-05,
|
| 463 |
+
"loss": 0.0581,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 3.6065573770491803,
|
| 468 |
+
"grad_norm": 0.5697107911109924,
|
| 469 |
+
"learning_rate": 9.042958964389427e-05,
|
| 470 |
+
"loss": 0.0578,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 3.66120218579235,
|
| 475 |
+
"grad_norm": 0.4503754675388336,
|
| 476 |
+
"learning_rate": 9.007215198239333e-05,
|
| 477 |
+
"loss": 0.0521,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 3.7158469945355193,
|
| 482 |
+
"grad_norm": 0.8456712365150452,
|
| 483 |
+
"learning_rate": 8.970889523063627e-05,
|
| 484 |
+
"loss": 0.0574,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 3.7704918032786887,
|
| 489 |
+
"grad_norm": 0.4836876392364502,
|
| 490 |
+
"learning_rate": 8.933987213906745e-05,
|
| 491 |
+
"loss": 0.0555,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 3.8251366120218577,
|
| 496 |
+
"grad_norm": 0.7264482378959656,
|
| 497 |
+
"learning_rate": 8.896513629549214e-05,
|
| 498 |
+
"loss": 0.0581,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 3.879781420765027,
|
| 503 |
+
"grad_norm": 0.43734481930732727,
|
| 504 |
+
"learning_rate": 8.858474211729469e-05,
|
| 505 |
+
"loss": 0.0522,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 3.9344262295081966,
|
| 510 |
+
"grad_norm": 0.6858192086219788,
|
| 511 |
+
"learning_rate": 8.81987448435363e-05,
|
| 512 |
+
"loss": 0.0526,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 3.989071038251366,
|
| 517 |
+
"grad_norm": 0.5418357253074646,
|
| 518 |
+
"learning_rate": 8.780720052693346e-05,
|
| 519 |
+
"loss": 0.0579,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 4.043715846994536,
|
| 524 |
+
"grad_norm": 0.6176689267158508,
|
| 525 |
+
"learning_rate": 8.741016602571826e-05,
|
| 526 |
+
"loss": 0.0617,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 4.098360655737705,
|
| 531 |
+
"grad_norm": 0.35425516963005066,
|
| 532 |
+
"learning_rate": 8.700769899538168e-05,
|
| 533 |
+
"loss": 0.0557,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 4.1530054644808745,
|
| 538 |
+
"grad_norm": 0.39824792742729187,
|
| 539 |
+
"learning_rate": 8.659985788030114e-05,
|
| 540 |
+
"loss": 0.0503,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 4.2076502732240435,
|
| 545 |
+
"grad_norm": 0.7548748254776001,
|
| 546 |
+
"learning_rate": 8.618670190525352e-05,
|
| 547 |
+
"loss": 0.0537,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 4.262295081967213,
|
| 552 |
+
"grad_norm": 0.4131401777267456,
|
| 553 |
+
"learning_rate": 8.57682910668147e-05,
|
| 554 |
+
"loss": 0.056,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 4.316939890710382,
|
| 559 |
+
"grad_norm": 0.44287577271461487,
|
| 560 |
+
"learning_rate": 8.534468612464725e-05,
|
| 561 |
+
"loss": 0.0491,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 4.371584699453552,
|
| 566 |
+
"grad_norm": 0.26551729440689087,
|
| 567 |
+
"learning_rate": 8.49159485926772e-05,
|
| 568 |
+
"loss": 0.0484,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 4.426229508196721,
|
| 573 |
+
"grad_norm": 0.9148038029670715,
|
| 574 |
+
"learning_rate": 8.448214073016116e-05,
|
| 575 |
+
"loss": 0.0489,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 4.48087431693989,
|
| 580 |
+
"grad_norm": 0.7359590530395508,
|
| 581 |
+
"learning_rate": 8.404332553264547e-05,
|
| 582 |
+
"loss": 0.0563,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 4.53551912568306,
|
| 587 |
+
"grad_norm": 0.512116551399231,
|
| 588 |
+
"learning_rate": 8.359956672281815e-05,
|
| 589 |
+
"loss": 0.0544,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 4.590163934426229,
|
| 594 |
+
"grad_norm": 0.455875039100647,
|
| 595 |
+
"learning_rate": 8.31509287412555e-05,
|
| 596 |
+
"loss": 0.0547,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 4.644808743169399,
|
| 601 |
+
"grad_norm": 0.5015522837638855,
|
| 602 |
+
"learning_rate": 8.269747673706422e-05,
|
| 603 |
+
"loss": 0.0537,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 4.699453551912568,
|
| 608 |
+
"grad_norm": 0.3985310196876526,
|
| 609 |
+
"learning_rate": 8.223927655842094e-05,
|
| 610 |
+
"loss": 0.0533,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 4.754098360655737,
|
| 615 |
+
"grad_norm": 0.4985620081424713,
|
| 616 |
+
"learning_rate": 8.177639474300983e-05,
|
| 617 |
+
"loss": 0.0468,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 4.808743169398907,
|
| 622 |
+
"grad_norm": 0.44651690125465393,
|
| 623 |
+
"learning_rate": 8.130889850836052e-05,
|
| 624 |
+
"loss": 0.0477,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 4.863387978142076,
|
| 629 |
+
"grad_norm": 0.6727082133293152,
|
| 630 |
+
"learning_rate": 8.083685574208694e-05,
|
| 631 |
+
"loss": 0.0497,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 4.918032786885246,
|
| 636 |
+
"grad_norm": 0.5853284001350403,
|
| 637 |
+
"learning_rate": 8.03603349920291e-05,
|
| 638 |
+
"loss": 0.0477,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 4.972677595628415,
|
| 643 |
+
"grad_norm": 0.41467201709747314,
|
| 644 |
+
"learning_rate": 7.987940545629877e-05,
|
| 645 |
+
"loss": 0.0466,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 5.027322404371585,
|
| 650 |
+
"grad_norm": 1.1400145292282104,
|
| 651 |
+
"learning_rate": 7.939413697323101e-05,
|
| 652 |
+
"loss": 0.0464,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 5.081967213114754,
|
| 657 |
+
"grad_norm": 0.6863431930541992,
|
| 658 |
+
"learning_rate": 7.890460001124242e-05,
|
| 659 |
+
"loss": 0.0506,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 5.136612021857924,
|
| 664 |
+
"grad_norm": 0.6057994365692139,
|
| 665 |
+
"learning_rate": 7.841086565859819e-05,
|
| 666 |
+
"loss": 0.0505,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 5.191256830601093,
|
| 671 |
+
"grad_norm": 0.6776031255722046,
|
| 672 |
+
"learning_rate": 7.791300561308888e-05,
|
| 673 |
+
"loss": 0.0498,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 5.245901639344262,
|
| 678 |
+
"grad_norm": 1.0112508535385132,
|
| 679 |
+
"learning_rate": 7.741109217161886e-05,
|
| 680 |
+
"loss": 0.0503,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 5.300546448087432,
|
| 685 |
+
"grad_norm": 0.4604274332523346,
|
| 686 |
+
"learning_rate": 7.690519821970768e-05,
|
| 687 |
+
"loss": 0.0495,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 5.355191256830601,
|
| 692 |
+
"grad_norm": 0.4461577832698822,
|
| 693 |
+
"learning_rate": 7.639539722090596e-05,
|
| 694 |
+
"loss": 0.0501,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 5.409836065573771,
|
| 699 |
+
"grad_norm": 0.39570051431655884,
|
| 700 |
+
"learning_rate": 7.588176320612741e-05,
|
| 701 |
+
"loss": 0.053,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 5.46448087431694,
|
| 706 |
+
"grad_norm": 0.3296869397163391,
|
| 707 |
+
"learning_rate": 7.536437076289833e-05,
|
| 708 |
+
"loss": 0.0487,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 5.51912568306011,
|
| 713 |
+
"grad_norm": 0.3206339180469513,
|
| 714 |
+
"learning_rate": 7.484329502452643e-05,
|
| 715 |
+
"loss": 0.0505,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 5.573770491803279,
|
| 720 |
+
"grad_norm": 0.4286837875843048,
|
| 721 |
+
"learning_rate": 7.431861165919027e-05,
|
| 722 |
+
"loss": 0.0437,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 5.628415300546449,
|
| 727 |
+
"grad_norm": 0.36809661984443665,
|
| 728 |
+
"learning_rate": 7.379039685895113e-05,
|
| 729 |
+
"loss": 0.043,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 5.683060109289618,
|
| 734 |
+
"grad_norm": 0.5219023823738098,
|
| 735 |
+
"learning_rate": 7.325872732868869e-05,
|
| 736 |
+
"loss": 0.0519,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 5.737704918032787,
|
| 741 |
+
"grad_norm": 0.2995985448360443,
|
| 742 |
+
"learning_rate": 7.27236802749624e-05,
|
| 743 |
+
"loss": 0.0451,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 5.7923497267759565,
|
| 748 |
+
"grad_norm": 0.4124583601951599,
|
| 749 |
+
"learning_rate": 7.218533339479976e-05,
|
| 750 |
+
"loss": 0.0435,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 5.8469945355191255,
|
| 755 |
+
"grad_norm": 0.39530396461486816,
|
| 756 |
+
"learning_rate": 7.164376486441373e-05,
|
| 757 |
+
"loss": 0.046,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 5.901639344262295,
|
| 762 |
+
"grad_norm": 0.7246448397636414,
|
| 763 |
+
"learning_rate": 7.109905332785014e-05,
|
| 764 |
+
"loss": 0.0523,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 5.956284153005464,
|
| 769 |
+
"grad_norm": 0.4755229651927948,
|
| 770 |
+
"learning_rate": 7.05512778855675e-05,
|
| 771 |
+
"loss": 0.0422,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 6.0109289617486334,
|
| 776 |
+
"grad_norm": 0.7326672077178955,
|
| 777 |
+
"learning_rate": 7.000051808295034e-05,
|
| 778 |
+
"loss": 0.0487,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 6.065573770491803,
|
| 783 |
+
"grad_norm": 0.44879159331321716,
|
| 784 |
+
"learning_rate": 6.944685389875804e-05,
|
| 785 |
+
"loss": 0.042,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 6.120218579234972,
|
| 790 |
+
"grad_norm": 0.7056254148483276,
|
| 791 |
+
"learning_rate": 6.889036573351067e-05,
|
| 792 |
+
"loss": 0.0449,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 6.174863387978142,
|
| 797 |
+
"grad_norm": 1.378929615020752,
|
| 798 |
+
"learning_rate": 6.833113439781355e-05,
|
| 799 |
+
"loss": 0.0474,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 6.229508196721311,
|
| 804 |
+
"grad_norm": 0.4310530126094818,
|
| 805 |
+
"learning_rate": 6.776924110062248e-05,
|
| 806 |
+
"loss": 0.042,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 6.284153005464481,
|
| 811 |
+
"grad_norm": 1.7358840703964233,
|
| 812 |
+
"learning_rate": 6.720476743745072e-05,
|
| 813 |
+
"loss": 0.0471,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 6.33879781420765,
|
| 818 |
+
"grad_norm": 0.5459834933280945,
|
| 819 |
+
"learning_rate": 6.66377953785203e-05,
|
| 820 |
+
"loss": 0.0468,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 6.39344262295082,
|
| 825 |
+
"grad_norm": 1.4089338779449463,
|
| 826 |
+
"learning_rate": 6.606840725685849e-05,
|
| 827 |
+
"loss": 0.0501,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 6.448087431693989,
|
| 832 |
+
"grad_norm": 0.5397401452064514,
|
| 833 |
+
"learning_rate": 6.549668575634193e-05,
|
| 834 |
+
"loss": 0.0437,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 6.502732240437158,
|
| 839 |
+
"grad_norm": 0.5156744122505188,
|
| 840 |
+
"learning_rate": 6.492271389968961e-05,
|
| 841 |
+
"loss": 0.0415,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 6.557377049180328,
|
| 846 |
+
"grad_norm": 0.47656911611557007,
|
| 847 |
+
"learning_rate": 6.434657503640666e-05,
|
| 848 |
+
"loss": 0.0408,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 6.612021857923497,
|
| 853 |
+
"grad_norm": 0.4555465281009674,
|
| 854 |
+
"learning_rate": 6.376835283068083e-05,
|
| 855 |
+
"loss": 0.0397,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 6.666666666666667,
|
| 860 |
+
"grad_norm": 0.5200902819633484,
|
| 861 |
+
"learning_rate": 6.318813124923303e-05,
|
| 862 |
+
"loss": 0.0431,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 6.721311475409836,
|
| 867 |
+
"grad_norm": 0.38336366415023804,
|
| 868 |
+
"learning_rate": 6.260599454912421e-05,
|
| 869 |
+
"loss": 0.0431,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 6.775956284153006,
|
| 874 |
+
"grad_norm": 0.6694422960281372,
|
| 875 |
+
"learning_rate": 6.202202726551985e-05,
|
| 876 |
+
"loss": 0.0459,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 6.830601092896175,
|
| 881 |
+
"grad_norm": 0.31609460711479187,
|
| 882 |
+
"learning_rate": 6.143631419941422e-05,
|
| 883 |
+
"loss": 0.0436,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 6.885245901639344,
|
| 888 |
+
"grad_norm": 0.43881747126579285,
|
| 889 |
+
"learning_rate": 6.08489404053159e-05,
|
| 890 |
+
"loss": 0.0442,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 6.939890710382514,
|
| 895 |
+
"grad_norm": 0.5141936540603638,
|
| 896 |
+
"learning_rate": 6.025999117889667e-05,
|
| 897 |
+
"loss": 0.0447,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 6.994535519125683,
|
| 902 |
+
"grad_norm": 0.6575435996055603,
|
| 903 |
+
"learning_rate": 5.966955204460517e-05,
|
| 904 |
+
"loss": 0.0348,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 7.049180327868853,
|
| 909 |
+
"grad_norm": 0.5895494818687439,
|
| 910 |
+
"learning_rate": 5.907770874324752e-05,
|
| 911 |
+
"loss": 0.0409,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 7.103825136612022,
|
| 916 |
+
"grad_norm": 0.28201115131378174,
|
| 917 |
+
"learning_rate": 5.8484547219536365e-05,
|
| 918 |
+
"loss": 0.0438,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 7.158469945355192,
|
| 923 |
+
"grad_norm": 0.408699631690979,
|
| 924 |
+
"learning_rate": 5.7890153609610454e-05,
|
| 925 |
+
"loss": 0.037,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 7.213114754098361,
|
| 930 |
+
"grad_norm": 0.6141229867935181,
|
| 931 |
+
"learning_rate": 5.7294614228526335e-05,
|
| 932 |
+
"loss": 0.0413,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 7.26775956284153,
|
| 937 |
+
"grad_norm": 0.64585942029953,
|
| 938 |
+
"learning_rate": 5.669801555772406e-05,
|
| 939 |
+
"loss": 0.0449,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 7.3224043715847,
|
| 944 |
+
"grad_norm": 0.5709479451179504,
|
| 945 |
+
"learning_rate": 5.6100444232468764e-05,
|
| 946 |
+
"loss": 0.0427,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 7.377049180327869,
|
| 951 |
+
"grad_norm": 0.5018634796142578,
|
| 952 |
+
"learning_rate": 5.5501987029269944e-05,
|
| 953 |
+
"loss": 0.0367,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 7.4316939890710385,
|
| 958 |
+
"grad_norm": 0.6372910141944885,
|
| 959 |
+
"learning_rate": 5.490273085328011e-05,
|
| 960 |
+
"loss": 0.0377,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 7.4863387978142075,
|
| 965 |
+
"grad_norm": 0.5395722389221191,
|
| 966 |
+
"learning_rate": 5.430276272567485e-05,
|
| 967 |
+
"loss": 0.0366,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 7.540983606557377,
|
| 972 |
+
"grad_norm": 0.2427688091993332,
|
| 973 |
+
"learning_rate": 5.3702169771016074e-05,
|
| 974 |
+
"loss": 0.0342,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 7.595628415300546,
|
| 979 |
+
"grad_norm": 0.5557045340538025,
|
| 980 |
+
"learning_rate": 5.3101039204600136e-05,
|
| 981 |
+
"loss": 0.0423,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 7.6502732240437155,
|
| 986 |
+
"grad_norm": 0.41853758692741394,
|
| 987 |
+
"learning_rate": 5.2499458319792816e-05,
|
| 988 |
+
"loss": 0.0408,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 7.704918032786885,
|
| 993 |
+
"grad_norm": 0.5503780245780945,
|
| 994 |
+
"learning_rate": 5.189751447535307e-05,
|
| 995 |
+
"loss": 0.0353,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 7.759562841530054,
|
| 1000 |
+
"grad_norm": 0.5195330381393433,
|
| 1001 |
+
"learning_rate": 5.129529508274712e-05,
|
| 1002 |
+
"loss": 0.0359,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 7.814207650273224,
|
| 1007 |
+
"grad_norm": 0.40231478214263916,
|
| 1008 |
+
"learning_rate": 5.069288759345502e-05,
|
| 1009 |
+
"loss": 0.0378,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 7.868852459016393,
|
| 1014 |
+
"grad_norm": 0.6481858491897583,
|
| 1015 |
+
"learning_rate": 5.009037948627134e-05,
|
| 1016 |
+
"loss": 0.0332,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 7.923497267759563,
|
| 1021 |
+
"grad_norm": 0.6040474772453308,
|
| 1022 |
+
"learning_rate": 4.948785825460191e-05,
|
| 1023 |
+
"loss": 0.0327,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 7.978142076502732,
|
| 1028 |
+
"grad_norm": 0.9101755619049072,
|
| 1029 |
+
"learning_rate": 4.888541139375841e-05,
|
| 1030 |
+
"loss": 0.0399,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 8.032786885245901,
|
| 1035 |
+
"grad_norm": 0.24122072756290436,
|
| 1036 |
+
"learning_rate": 4.8283126388252794e-05,
|
| 1037 |
+
"loss": 0.0354,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 8.087431693989071,
|
| 1042 |
+
"grad_norm": 0.47983938455581665,
|
| 1043 |
+
"learning_rate": 4.768109069909307e-05,
|
| 1044 |
+
"loss": 0.0357,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 8.142076502732241,
|
| 1049 |
+
"grad_norm": 1.1941678524017334,
|
| 1050 |
+
"learning_rate": 4.7079391751082773e-05,
|
| 1051 |
+
"loss": 0.0375,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 8.19672131147541,
|
| 1056 |
+
"grad_norm": 0.424197256565094,
|
| 1057 |
+
"learning_rate": 4.64781169201254e-05,
|
| 1058 |
+
"loss": 0.032,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 8.251366120218579,
|
| 1063 |
+
"grad_norm": 0.4514443874359131,
|
| 1064 |
+
"learning_rate": 4.5877353520536177e-05,
|
| 1065 |
+
"loss": 0.0407,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 8.306010928961749,
|
| 1070 |
+
"grad_norm": 0.33258920907974243,
|
| 1071 |
+
"learning_rate": 4.5277188792362705e-05,
|
| 1072 |
+
"loss": 0.0359,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 8.360655737704919,
|
| 1077 |
+
"grad_norm": 0.7045204043388367,
|
| 1078 |
+
"learning_rate": 4.467770988871621e-05,
|
| 1079 |
+
"loss": 0.0406,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 8.415300546448087,
|
| 1084 |
+
"grad_norm": 1.1585627794265747,
|
| 1085 |
+
"learning_rate": 4.407900386311582e-05,
|
| 1086 |
+
"loss": 0.0357,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 8.469945355191257,
|
| 1091 |
+
"grad_norm": 0.3667105436325073,
|
| 1092 |
+
"learning_rate": 4.348115765684687e-05,
|
| 1093 |
+
"loss": 0.032,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 8.524590163934427,
|
| 1098 |
+
"grad_norm": 0.6133992671966553,
|
| 1099 |
+
"learning_rate": 4.288425808633575e-05,
|
| 1100 |
+
"loss": 0.0357,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 8.579234972677595,
|
| 1105 |
+
"grad_norm": 0.6548840403556824,
|
| 1106 |
+
"learning_rate": 4.2288391830542885e-05,
|
| 1107 |
+
"loss": 0.0382,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 8.633879781420765,
|
| 1112 |
+
"grad_norm": 0.6242523789405823,
|
| 1113 |
+
"learning_rate": 4.1693645418375526e-05,
|
| 1114 |
+
"loss": 0.0345,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 8.688524590163935,
|
| 1119 |
+
"grad_norm": 0.7505032420158386,
|
| 1120 |
+
"learning_rate": 4.11001052161225e-05,
|
| 1121 |
+
"loss": 0.0335,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 8.743169398907105,
|
| 1126 |
+
"grad_norm": 1.287976861000061,
|
| 1127 |
+
"learning_rate": 4.050785741491248e-05,
|
| 1128 |
+
"loss": 0.0375,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 8.797814207650273,
|
| 1133 |
+
"grad_norm": 0.4427262842655182,
|
| 1134 |
+
"learning_rate": 3.9916988018197694e-05,
|
| 1135 |
+
"loss": 0.0359,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 8.852459016393443,
|
| 1140 |
+
"grad_norm": 1.1378114223480225,
|
| 1141 |
+
"learning_rate": 3.9327582829264964e-05,
|
| 1142 |
+
"loss": 0.0337,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 8.907103825136613,
|
| 1147 |
+
"grad_norm": 0.34418851137161255,
|
| 1148 |
+
"learning_rate": 3.8739727438775676e-05,
|
| 1149 |
+
"loss": 0.0302,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 8.96174863387978,
|
| 1154 |
+
"grad_norm": 0.9304275512695312,
|
| 1155 |
+
"learning_rate": 3.815350721233674e-05,
|
| 1156 |
+
"loss": 0.0371,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 9.01639344262295,
|
| 1161 |
+
"grad_norm": 0.4125829339027405,
|
| 1162 |
+
"learning_rate": 3.756900727810423e-05,
|
| 1163 |
+
"loss": 0.0335,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 9.07103825136612,
|
| 1168 |
+
"grad_norm": 0.9346985816955566,
|
| 1169 |
+
"learning_rate": 3.698631251442144e-05,
|
| 1170 |
+
"loss": 0.0347,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 9.12568306010929,
|
| 1175 |
+
"grad_norm": 0.31927409768104553,
|
| 1176 |
+
"learning_rate": 3.640550753749321e-05,
|
| 1177 |
+
"loss": 0.0348,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 9.180327868852459,
|
| 1182 |
+
"grad_norm": 0.35485345125198364,
|
| 1183 |
+
"learning_rate": 3.582667668909846e-05,
|
| 1184 |
+
"loss": 0.0317,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 9.234972677595628,
|
| 1189 |
+
"grad_norm": 0.5352278351783752,
|
| 1190 |
+
"learning_rate": 3.524990402434234e-05,
|
| 1191 |
+
"loss": 0.0336,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 9.289617486338798,
|
| 1196 |
+
"grad_norm": 0.4585462808609009,
|
| 1197 |
+
"learning_rate": 3.467527329945026e-05,
|
| 1198 |
+
"loss": 0.0277,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 9.344262295081966,
|
| 1203 |
+
"grad_norm": 0.6951584219932556,
|
| 1204 |
+
"learning_rate": 3.410286795960514e-05,
|
| 1205 |
+
"loss": 0.0319,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 9.398907103825136,
|
| 1210 |
+
"grad_norm": 0.36788222193717957,
|
| 1211 |
+
"learning_rate": 3.353277112682991e-05,
|
| 1212 |
+
"loss": 0.0302,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 9.453551912568306,
|
| 1217 |
+
"grad_norm": 0.23181340098381042,
|
| 1218 |
+
"learning_rate": 3.2965065587916886e-05,
|
| 1219 |
+
"loss": 0.0266,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 9.508196721311476,
|
| 1224 |
+
"grad_norm": 0.28742238879203796,
|
| 1225 |
+
"learning_rate": 3.239983378240593e-05,
|
| 1226 |
+
"loss": 0.0372,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 9.562841530054644,
|
| 1231 |
+
"grad_norm": 0.509660542011261,
|
| 1232 |
+
"learning_rate": 3.183715779061288e-05,
|
| 1233 |
+
"loss": 0.0348,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 9.617486338797814,
|
| 1238 |
+
"grad_norm": 0.4197666049003601,
|
| 1239 |
+
"learning_rate": 3.1277119321710314e-05,
|
| 1240 |
+
"loss": 0.0339,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 9.672131147540984,
|
| 1245 |
+
"grad_norm": 0.7674693465232849,
|
| 1246 |
+
"learning_rate": 3.0719799701862065e-05,
|
| 1247 |
+
"loss": 0.0359,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 9.726775956284152,
|
| 1252 |
+
"grad_norm": 1.1442123651504517,
|
| 1253 |
+
"learning_rate": 3.0165279862413432e-05,
|
| 1254 |
+
"loss": 0.0318,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 9.781420765027322,
|
| 1259 |
+
"grad_norm": 0.646183967590332,
|
| 1260 |
+
"learning_rate": 2.9613640328138748e-05,
|
| 1261 |
+
"loss": 0.0328,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 9.836065573770492,
|
| 1266 |
+
"grad_norm": 1.4289329051971436,
|
| 1267 |
+
"learning_rate": 2.906496120554789e-05,
|
| 1268 |
+
"loss": 0.0306,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 9.890710382513662,
|
| 1273 |
+
"grad_norm": 0.2373889982700348,
|
| 1274 |
+
"learning_rate": 2.8519322171253602e-05,
|
| 1275 |
+
"loss": 0.0297,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 9.94535519125683,
|
| 1280 |
+
"grad_norm": 0.5010556578636169,
|
| 1281 |
+
"learning_rate": 2.7976802460401287e-05,
|
| 1282 |
+
"loss": 0.0316,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 10.0,
|
| 1287 |
+
"grad_norm": 0.4153306782245636,
|
| 1288 |
+
"learning_rate": 2.743748085516275e-05,
|
| 1289 |
+
"loss": 0.0311,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 10.05464480874317,
|
| 1294 |
+
"grad_norm": 0.4151310622692108,
|
| 1295 |
+
"learning_rate": 2.6901435673295917e-05,
|
| 1296 |
+
"loss": 0.0262,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 10.109289617486338,
|
| 1301 |
+
"grad_norm": 0.2862219214439392,
|
| 1302 |
+
"learning_rate": 2.6368744756771837e-05,
|
| 1303 |
+
"loss": 0.0297,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 10.163934426229508,
|
| 1308 |
+
"grad_norm": 0.236620232462883,
|
| 1309 |
+
"learning_rate": 2.5839485460470862e-05,
|
| 1310 |
+
"loss": 0.0277,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 10.218579234972678,
|
| 1315 |
+
"grad_norm": 0.4799292981624603,
|
| 1316 |
+
"learning_rate": 2.53137346409495e-05,
|
| 1317 |
+
"loss": 0.031,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 10.273224043715848,
|
| 1322 |
+
"grad_norm": 0.3376322090625763,
|
| 1323 |
+
"learning_rate": 2.4791568645279746e-05,
|
| 1324 |
+
"loss": 0.0257,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 10.327868852459016,
|
| 1329 |
+
"grad_norm": 0.31595292687416077,
|
| 1330 |
+
"learning_rate": 2.4273063299962212e-05,
|
| 1331 |
+
"loss": 0.0282,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 10.382513661202186,
|
| 1336 |
+
"grad_norm": 0.6953328251838684,
|
| 1337 |
+
"learning_rate": 2.375829389991503e-05,
|
| 1338 |
+
"loss": 0.0365,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 10.437158469945356,
|
| 1343 |
+
"grad_norm": 0.48706644773483276,
|
| 1344 |
+
"learning_rate": 2.3247335197539843e-05,
|
| 1345 |
+
"loss": 0.0291,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 10.491803278688524,
|
| 1350 |
+
"grad_norm": 2.7361419200897217,
|
| 1351 |
+
"learning_rate": 2.2740261391866637e-05,
|
| 1352 |
+
"loss": 0.031,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 10.546448087431694,
|
| 1357 |
+
"grad_norm": 0.3964270055294037,
|
| 1358 |
+
"learning_rate": 2.2237146117778857e-05,
|
| 1359 |
+
"loss": 0.0332,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 10.601092896174864,
|
| 1364 |
+
"grad_norm": 0.743932843208313,
|
| 1365 |
+
"learning_rate": 2.17380624353206e-05,
|
| 1366 |
+
"loss": 0.0316,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 10.655737704918034,
|
| 1371 |
+
"grad_norm": 0.32625240087509155,
|
| 1372 |
+
"learning_rate": 2.1243082819087044e-05,
|
| 1373 |
+
"loss": 0.0273,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 10.710382513661202,
|
| 1378 |
+
"grad_norm": 0.45872801542282104,
|
| 1379 |
+
"learning_rate": 2.075227914770011e-05,
|
| 1380 |
+
"loss": 0.0275,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 10.765027322404372,
|
| 1385 |
+
"grad_norm": 0.5026045441627502,
|
| 1386 |
+
"learning_rate": 2.026572269337056e-05,
|
| 1387 |
+
"loss": 0.0284,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 10.819672131147541,
|
| 1392 |
+
"grad_norm": 0.544503927230835,
|
| 1393 |
+
"learning_rate": 1.978348411154816e-05,
|
| 1394 |
+
"loss": 0.0309,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 10.87431693989071,
|
| 1399 |
+
"grad_norm": 0.44638386368751526,
|
| 1400 |
+
"learning_rate": 1.9305633430661522e-05,
|
| 1401 |
+
"loss": 0.0245,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 10.92896174863388,
|
| 1406 |
+
"grad_norm": 0.8865587711334229,
|
| 1407 |
+
"learning_rate": 1.8832240041948773e-05,
|
| 1408 |
+
"loss": 0.0292,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 10.98360655737705,
|
| 1413 |
+
"grad_norm": 0.3047027587890625,
|
| 1414 |
+
"learning_rate": 1.8363372689380943e-05,
|
| 1415 |
+
"loss": 0.0299,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 11.03825136612022,
|
| 1420 |
+
"grad_norm": 0.42466291785240173,
|
| 1421 |
+
"learning_rate": 1.7899099459679315e-05,
|
| 1422 |
+
"loss": 0.0309,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 11.092896174863387,
|
| 1427 |
+
"grad_norm": 0.8315626382827759,
|
| 1428 |
+
"learning_rate": 1.743948777242814e-05,
|
| 1429 |
+
"loss": 0.0321,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 11.147540983606557,
|
| 1434 |
+
"grad_norm": 0.58014976978302,
|
| 1435 |
+
"learning_rate": 1.6984604370284313e-05,
|
| 1436 |
+
"loss": 0.0277,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 11.202185792349727,
|
| 1441 |
+
"grad_norm": 0.3877146542072296,
|
| 1442 |
+
"learning_rate": 1.6534515309285333e-05,
|
| 1443 |
+
"loss": 0.0311,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 11.256830601092895,
|
| 1448 |
+
"grad_norm": 0.5980963706970215,
|
| 1449 |
+
"learning_rate": 1.608928594925689e-05,
|
| 1450 |
+
"loss": 0.0277,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 11.311475409836065,
|
| 1455 |
+
"grad_norm": 0.2995881140232086,
|
| 1456 |
+
"learning_rate": 1.564898094432168e-05,
|
| 1457 |
+
"loss": 0.0294,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 11.366120218579235,
|
| 1462 |
+
"grad_norm": 0.3173847794532776,
|
| 1463 |
+
"learning_rate": 1.5213664233510595e-05,
|
| 1464 |
+
"loss": 0.0248,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 11.420765027322405,
|
| 1469 |
+
"grad_norm": 0.4316744804382324,
|
| 1470 |
+
"learning_rate": 1.4783399031477835e-05,
|
| 1471 |
+
"loss": 0.0302,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 11.475409836065573,
|
| 1476 |
+
"grad_norm": 0.4270103871822357,
|
| 1477 |
+
"learning_rate": 1.4358247819321119e-05,
|
| 1478 |
+
"loss": 0.0288,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 11.530054644808743,
|
| 1483 |
+
"grad_norm": 0.5547152757644653,
|
| 1484 |
+
"learning_rate": 1.3938272335508561e-05,
|
| 1485 |
+
"loss": 0.0281,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 11.584699453551913,
|
| 1490 |
+
"grad_norm": 0.4367380738258362,
|
| 1491 |
+
"learning_rate": 1.3523533566913194e-05,
|
| 1492 |
+
"loss": 0.0253,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 11.639344262295083,
|
| 1497 |
+
"grad_norm": 0.36034998297691345,
|
| 1498 |
+
"learning_rate": 1.3114091739956796e-05,
|
| 1499 |
+
"loss": 0.0311,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 11.693989071038251,
|
| 1504 |
+
"grad_norm": 0.32479503750801086,
|
| 1505 |
+
"learning_rate": 1.2710006311864104e-05,
|
| 1506 |
+
"loss": 0.0279,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 11.748633879781421,
|
| 1511 |
+
"grad_norm": 0.7293975353240967,
|
| 1512 |
+
"learning_rate": 1.231133596202868e-05,
|
| 1513 |
+
"loss": 0.0257,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 11.80327868852459,
|
| 1518 |
+
"grad_norm": 0.18685133755207062,
|
| 1519 |
+
"learning_rate": 1.1918138583491784e-05,
|
| 1520 |
+
"loss": 0.0231,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 11.857923497267759,
|
| 1525 |
+
"grad_norm": 0.24629777669906616,
|
| 1526 |
+
"learning_rate": 1.1530471274535414e-05,
|
| 1527 |
+
"loss": 0.0257,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 11.912568306010929,
|
| 1532 |
+
"grad_norm": 0.5375162959098816,
|
| 1533 |
+
"learning_rate": 1.1148390330390767e-05,
|
| 1534 |
+
"loss": 0.0274,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 11.967213114754099,
|
| 1539 |
+
"grad_norm": 0.6662833094596863,
|
| 1540 |
+
"learning_rate": 1.0771951235063288e-05,
|
| 1541 |
+
"loss": 0.0299,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 12.021857923497267,
|
| 1546 |
+
"grad_norm": 0.3094324469566345,
|
| 1547 |
+
"learning_rate": 1.0401208653275556e-05,
|
| 1548 |
+
"loss": 0.0294,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 12.076502732240437,
|
| 1553 |
+
"grad_norm": 0.37933555245399475,
|
| 1554 |
+
"learning_rate": 1.0036216422529171e-05,
|
| 1555 |
+
"loss": 0.0298,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 12.131147540983607,
|
| 1560 |
+
"grad_norm": 0.42512938380241394,
|
| 1561 |
+
"learning_rate": 9.677027545286638e-06,
|
| 1562 |
+
"loss": 0.0291,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 12.185792349726777,
|
| 1567 |
+
"grad_norm": 0.4736573398113251,
|
| 1568 |
+
"learning_rate": 9.323694181274645e-06,
|
| 1569 |
+
"loss": 0.0288,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 12.240437158469945,
|
| 1574 |
+
"grad_norm": 0.6718573570251465,
|
| 1575 |
+
"learning_rate": 8.976267639909697e-06,
|
| 1576 |
+
"loss": 0.0279,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 12.295081967213115,
|
| 1581 |
+
"grad_norm": 0.2468743771314621,
|
| 1582 |
+
"learning_rate": 8.634798372847148e-06,
|
| 1583 |
+
"loss": 0.0288,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 12.349726775956285,
|
| 1588 |
+
"grad_norm": 0.4593784809112549,
|
| 1589 |
+
"learning_rate": 8.299335966654882e-06,
|
| 1590 |
+
"loss": 0.0281,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 12.404371584699454,
|
| 1595 |
+
"grad_norm": 0.19763712584972382,
|
| 1596 |
+
"learning_rate": 7.969929135612603e-06,
|
| 1597 |
+
"loss": 0.0226,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 12.459016393442623,
|
| 1602 |
+
"grad_norm": 1.0427390336990356,
|
| 1603 |
+
"learning_rate": 7.646625714637712e-06,
|
| 1604 |
+
"loss": 0.026,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 12.513661202185792,
|
| 1609 |
+
"grad_norm": 1.0317535400390625,
|
| 1610 |
+
"learning_rate": 7.329472652339009e-06,
|
| 1611 |
+
"loss": 0.0326,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 12.568306010928962,
|
| 1616 |
+
"grad_norm": 0.7976316809654236,
|
| 1617 |
+
"learning_rate": 7.018516004198994e-06,
|
| 1618 |
+
"loss": 0.0237,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 12.62295081967213,
|
| 1623 |
+
"grad_norm": 0.3223474621772766,
|
| 1624 |
+
"learning_rate": 6.713800925885905e-06,
|
| 1625 |
+
"loss": 0.0274,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 12.6775956284153,
|
| 1630 |
+
"grad_norm": 0.7897388935089111,
|
| 1631 |
+
"learning_rate": 6.4153716666964906e-06,
|
| 1632 |
+
"loss": 0.028,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 12.73224043715847,
|
| 1637 |
+
"grad_norm": 0.6605315804481506,
|
| 1638 |
+
"learning_rate": 6.123271563130234e-06,
|
| 1639 |
+
"loss": 0.0278,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 12.78688524590164,
|
| 1644 |
+
"grad_norm": 0.22467422485351562,
|
| 1645 |
+
"learning_rate": 5.837543032596293e-06,
|
| 1646 |
+
"loss": 0.0254,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 12.841530054644808,
|
| 1651 |
+
"grad_norm": 0.4244163930416107,
|
| 1652 |
+
"learning_rate": 5.558227567253832e-06,
|
| 1653 |
+
"loss": 0.0254,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 12.896174863387978,
|
| 1658 |
+
"grad_norm": 0.5720034241676331,
|
| 1659 |
+
"learning_rate": 5.285365727986707e-06,
|
| 1660 |
+
"loss": 0.025,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 12.950819672131148,
|
| 1665 |
+
"grad_norm": 0.4090539515018463,
|
| 1666 |
+
"learning_rate": 5.018997138513421e-06,
|
| 1667 |
+
"loss": 0.0281,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 13.005464480874316,
|
| 1672 |
+
"grad_norm": 0.9096952080726624,
|
| 1673 |
+
"learning_rate": 4.759160479633174e-06,
|
| 1674 |
+
"loss": 0.0293,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 13.060109289617486,
|
| 1679 |
+
"grad_norm": 0.2591124176979065,
|
| 1680 |
+
"learning_rate": 4.505893483608759e-06,
|
| 1681 |
+
"loss": 0.0251,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 13.114754098360656,
|
| 1686 |
+
"grad_norm": 0.2549198567867279,
|
| 1687 |
+
"learning_rate": 4.259232928687318e-06,
|
| 1688 |
+
"loss": 0.0281,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 13.169398907103826,
|
| 1693 |
+
"grad_norm": 0.39519423246383667,
|
| 1694 |
+
"learning_rate": 4.019214633759532e-06,
|
| 1695 |
+
"loss": 0.0288,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 13.224043715846994,
|
| 1700 |
+
"grad_norm": 0.4667559266090393,
|
| 1701 |
+
"learning_rate": 3.7858734531582175e-06,
|
| 1702 |
+
"loss": 0.0224,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 13.278688524590164,
|
| 1707 |
+
"grad_norm": 0.553810715675354,
|
| 1708 |
+
"learning_rate": 3.5592432715968902e-06,
|
| 1709 |
+
"loss": 0.0293,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 13.333333333333334,
|
| 1714 |
+
"grad_norm": 0.6030510663986206,
|
| 1715 |
+
"learning_rate": 3.3393569992492434e-06,
|
| 1716 |
+
"loss": 0.023,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 13.387978142076502,
|
| 1721 |
+
"grad_norm": 0.41252389550209045,
|
| 1722 |
+
"learning_rate": 3.1262465669700204e-06,
|
| 1723 |
+
"loss": 0.0261,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 13.442622950819672,
|
| 1728 |
+
"grad_norm": 0.2342725694179535,
|
| 1729 |
+
"learning_rate": 2.9199429216582195e-06,
|
| 1730 |
+
"loss": 0.0294,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 13.497267759562842,
|
| 1735 |
+
"grad_norm": 0.9791005849838257,
|
| 1736 |
+
"learning_rate": 2.7204760217631074e-06,
|
| 1737 |
+
"loss": 0.025,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 13.551912568306012,
|
| 1742 |
+
"grad_norm": 0.7885070443153381,
|
| 1743 |
+
"learning_rate": 2.5278748329338074e-06,
|
| 1744 |
+
"loss": 0.0228,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 13.60655737704918,
|
| 1749 |
+
"grad_norm": 0.2128724902868271,
|
| 1750 |
+
"learning_rate": 2.3421673238130215e-06,
|
| 1751 |
+
"loss": 0.0224,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 13.66120218579235,
|
| 1756 |
+
"grad_norm": 0.6793450713157654,
|
| 1757 |
+
"learning_rate": 2.163380461975573e-06,
|
| 1758 |
+
"loss": 0.0306,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 13.71584699453552,
|
| 1763 |
+
"grad_norm": 0.3011026978492737,
|
| 1764 |
+
"learning_rate": 1.991540210012288e-06,
|
| 1765 |
+
"loss": 0.0231,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 13.770491803278688,
|
| 1770 |
+
"grad_norm": 0.22343051433563232,
|
| 1771 |
+
"learning_rate": 1.826671521759843e-06,
|
| 1772 |
+
"loss": 0.0237,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 13.825136612021858,
|
| 1777 |
+
"grad_norm": 0.7726356387138367,
|
| 1778 |
+
"learning_rate": 1.6687983386770744e-06,
|
| 1779 |
+
"loss": 0.0247,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 13.879781420765028,
|
| 1784 |
+
"grad_norm": 0.2594599723815918,
|
| 1785 |
+
"learning_rate": 1.5179435863683368e-06,
|
| 1786 |
+
"loss": 0.0228,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 13.934426229508198,
|
| 1791 |
+
"grad_norm": 0.6315459609031677,
|
| 1792 |
+
"learning_rate": 1.374129171254307e-06,
|
| 1793 |
+
"loss": 0.0269,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 13.989071038251366,
|
| 1798 |
+
"grad_norm": 0.5242689847946167,
|
| 1799 |
+
"learning_rate": 1.2373759773908754e-06,
|
| 1800 |
+
"loss": 0.0251,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 14.043715846994536,
|
| 1805 |
+
"grad_norm": 0.37615323066711426,
|
| 1806 |
+
"learning_rate": 1.1077038634364601e-06,
|
| 1807 |
+
"loss": 0.0245,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 14.098360655737705,
|
| 1812 |
+
"grad_norm": 0.31767016649246216,
|
| 1813 |
+
"learning_rate": 9.851316597681958e-07,
|
| 1814 |
+
"loss": 0.0273,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 14.153005464480874,
|
| 1819 |
+
"grad_norm": 0.30916398763656616,
|
| 1820 |
+
"learning_rate": 8.696771657475067e-07,
|
| 1821 |
+
"loss": 0.0248,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 14.207650273224044,
|
| 1826 |
+
"grad_norm": 0.6139862537384033,
|
| 1827 |
+
"learning_rate": 7.613571471353497e-07,
|
| 1828 |
+
"loss": 0.0283,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 14.262295081967213,
|
| 1833 |
+
"grad_norm": 0.4068570137023926,
|
| 1834 |
+
"learning_rate": 6.601873336575748e-07,
|
| 1835 |
+
"loss": 0.0234,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 14.316939890710383,
|
| 1840 |
+
"grad_norm": 0.30164000391960144,
|
| 1841 |
+
"learning_rate": 5.66182416720723e-07,
|
| 1842 |
+
"loss": 0.0281,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 14.371584699453551,
|
| 1847 |
+
"grad_norm": 0.34394413232803345,
|
| 1848 |
+
"learning_rate": 4.793560472786173e-07,
|
| 1849 |
+
"loss": 0.0243,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 14.426229508196721,
|
| 1854 |
+
"grad_norm": 0.9818114042282104,
|
| 1855 |
+
"learning_rate": 3.9972083385002026e-07,
|
| 1856 |
+
"loss": 0.0286,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 14.480874316939891,
|
| 1861 |
+
"grad_norm": 0.3286588788032532,
|
| 1862 |
+
"learning_rate": 3.272883406877092e-07,
|
| 1863 |
+
"loss": 0.0248,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 14.53551912568306,
|
| 1868 |
+
"grad_norm": 0.38933223485946655,
|
| 1869 |
+
"learning_rate": 2.620690860991426e-07,
|
| 1870 |
+
"loss": 0.0289,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 14.59016393442623,
|
| 1875 |
+
"grad_norm": 0.3602127730846405,
|
| 1876 |
+
"learning_rate": 2.040725409190536e-07,
|
| 1877 |
+
"loss": 0.0222,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 14.6448087431694,
|
| 1882 |
+
"grad_norm": 0.21410511434078217,
|
| 1883 |
+
"learning_rate": 1.5330712713413907e-07,
|
| 1884 |
+
"loss": 0.0251,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 14.699453551912569,
|
| 1889 |
+
"grad_norm": 0.3144880533218384,
|
| 1890 |
+
"learning_rate": 1.0978021666005478e-07,
|
| 1891 |
+
"loss": 0.0248,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 14.754098360655737,
|
| 1896 |
+
"grad_norm": 0.5586285591125488,
|
| 1897 |
+
"learning_rate": 7.34981302709159e-08,
|
| 1898 |
+
"loss": 0.027,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 14.808743169398907,
|
| 1903 |
+
"grad_norm": 0.17363548278808594,
|
| 1904 |
+
"learning_rate": 4.4466136681398096e-08,
|
| 1905 |
+
"loss": 0.0242,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 14.863387978142077,
|
| 1910 |
+
"grad_norm": 0.27791473269462585,
|
| 1911 |
+
"learning_rate": 2.268845178165502e-08,
|
| 1912 |
+
"loss": 0.0234,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 14.918032786885245,
|
| 1917 |
+
"grad_norm": 0.29759353399276733,
|
| 1918 |
+
"learning_rate": 8.16823802510247e-09,
|
| 1919 |
+
"loss": 0.0295,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 14.972677595628415,
|
| 1924 |
+
"grad_norm": 0.24673596024513245,
|
| 1925 |
+
"learning_rate": 9.076039691913174e-10,
|
| 1926 |
+
"loss": 0.0277,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 15.0,
|
| 1931 |
+
"step": 2745,
|
| 1932 |
+
"total_flos": 2.913783625748952e+17,
|
| 1933 |
+
"train_loss": 0.06019136120984246,
|
| 1934 |
+
"train_runtime": 2906.6943,
|
| 1935 |
+
"train_samples_per_second": 46.264,
|
| 1936 |
+
"train_steps_per_second": 0.944
|
| 1937 |
+
}
|
| 1938 |
+
],
|
| 1939 |
+
"logging_steps": 10,
|
| 1940 |
+
"max_steps": 2745,
|
| 1941 |
+
"num_input_tokens_seen": 0,
|
| 1942 |
+
"num_train_epochs": 15,
|
| 1943 |
+
"save_steps": 10000,
|
| 1944 |
+
"stateful_callbacks": {
|
| 1945 |
+
"TrainerControl": {
|
| 1946 |
+
"args": {
|
| 1947 |
+
"should_epoch_stop": false,
|
| 1948 |
+
"should_evaluate": false,
|
| 1949 |
+
"should_log": false,
|
| 1950 |
+
"should_save": true,
|
| 1951 |
+
"should_training_stop": true
|
| 1952 |
+
},
|
| 1953 |
+
"attributes": {}
|
| 1954 |
+
}
|
| 1955 |
+
},
|
| 1956 |
+
"total_flos": 2.913783625748952e+17,
|
| 1957 |
+
"train_batch_size": 49,
|
| 1958 |
+
"trial_name": null,
|
| 1959 |
+
"trial_params": null
|
| 1960 |
+
}
|