Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1841 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1841 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.31899641577061,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2600,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.04,
|
| 13 |
+
"grad_norm": 240.98789978027344,
|
| 14 |
+
"learning_rate": 9.978494623655915e-06,
|
| 15 |
+
"loss": 4.1336,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.07,
|
| 20 |
+
"grad_norm": NaN,
|
| 21 |
+
"learning_rate": 9.953405017921148e-06,
|
| 22 |
+
"loss": 3.7709,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.11,
|
| 27 |
+
"grad_norm": 414.7696533203125,
|
| 28 |
+
"learning_rate": 9.917562724014338e-06,
|
| 29 |
+
"loss": 3.5379,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.14,
|
| 34 |
+
"grad_norm": 160.37644958496094,
|
| 35 |
+
"learning_rate": 9.881720430107527e-06,
|
| 36 |
+
"loss": 3.2512,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.18,
|
| 41 |
+
"grad_norm": 1477.7020263671875,
|
| 42 |
+
"learning_rate": 9.845878136200718e-06,
|
| 43 |
+
"loss": 2.8186,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.22,
|
| 48 |
+
"grad_norm": 444.38775634765625,
|
| 49 |
+
"learning_rate": 9.810035842293908e-06,
|
| 50 |
+
"loss": 2.5583,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.25,
|
| 55 |
+
"grad_norm": 119.81842041015625,
|
| 56 |
+
"learning_rate": 9.774193548387097e-06,
|
| 57 |
+
"loss": 2.4916,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.29,
|
| 62 |
+
"grad_norm": 175.25006103515625,
|
| 63 |
+
"learning_rate": 9.74193548387097e-06,
|
| 64 |
+
"loss": 2.4299,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.32,
|
| 69 |
+
"grad_norm": 116.39012145996094,
|
| 70 |
+
"learning_rate": 9.706093189964158e-06,
|
| 71 |
+
"loss": 2.2983,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.36,
|
| 76 |
+
"grad_norm": 313.2846984863281,
|
| 77 |
+
"learning_rate": 9.670250896057349e-06,
|
| 78 |
+
"loss": 2.233,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.39,
|
| 83 |
+
"grad_norm": 228.02125549316406,
|
| 84 |
+
"learning_rate": 9.634408602150539e-06,
|
| 85 |
+
"loss": 2.0787,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.43,
|
| 90 |
+
"grad_norm": 82.76283264160156,
|
| 91 |
+
"learning_rate": 9.598566308243728e-06,
|
| 92 |
+
"loss": 2.2126,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.47,
|
| 97 |
+
"grad_norm": 94.6029052734375,
|
| 98 |
+
"learning_rate": 9.562724014336918e-06,
|
| 99 |
+
"loss": 2.0342,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.5,
|
| 104 |
+
"grad_norm": 259.7580871582031,
|
| 105 |
+
"learning_rate": 9.526881720430107e-06,
|
| 106 |
+
"loss": 1.8917,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.54,
|
| 111 |
+
"grad_norm": 453.753173828125,
|
| 112 |
+
"learning_rate": 9.491039426523298e-06,
|
| 113 |
+
"loss": 2.0701,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.57,
|
| 118 |
+
"grad_norm": 131.5532989501953,
|
| 119 |
+
"learning_rate": 9.455197132616488e-06,
|
| 120 |
+
"loss": 1.8425,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.61,
|
| 125 |
+
"grad_norm": 528.49560546875,
|
| 126 |
+
"learning_rate": 9.419354838709677e-06,
|
| 127 |
+
"loss": 1.7829,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.65,
|
| 132 |
+
"grad_norm": 510.9673156738281,
|
| 133 |
+
"learning_rate": 9.38351254480287e-06,
|
| 134 |
+
"loss": 1.8383,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.68,
|
| 139 |
+
"grad_norm": 140.07981872558594,
|
| 140 |
+
"learning_rate": 9.347670250896058e-06,
|
| 141 |
+
"loss": 2.0396,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.72,
|
| 146 |
+
"grad_norm": 284.4012145996094,
|
| 147 |
+
"learning_rate": 9.311827956989249e-06,
|
| 148 |
+
"loss": 1.8723,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.75,
|
| 153 |
+
"grad_norm": 105.90425109863281,
|
| 154 |
+
"learning_rate": 9.27598566308244e-06,
|
| 155 |
+
"loss": 1.8539,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.79,
|
| 160 |
+
"grad_norm": 77.53498077392578,
|
| 161 |
+
"learning_rate": 9.240143369175628e-06,
|
| 162 |
+
"loss": 1.8487,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.82,
|
| 167 |
+
"grad_norm": 203.25697326660156,
|
| 168 |
+
"learning_rate": 9.204301075268819e-06,
|
| 169 |
+
"loss": 1.7915,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.86,
|
| 174 |
+
"grad_norm": 167.5474853515625,
|
| 175 |
+
"learning_rate": 9.168458781362007e-06,
|
| 176 |
+
"loss": 1.6751,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.9,
|
| 181 |
+
"grad_norm": 141.84117126464844,
|
| 182 |
+
"learning_rate": 9.132616487455198e-06,
|
| 183 |
+
"loss": 1.7533,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.93,
|
| 188 |
+
"grad_norm": 170.94772338867188,
|
| 189 |
+
"learning_rate": 9.096774193548388e-06,
|
| 190 |
+
"loss": 1.6008,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.97,
|
| 195 |
+
"grad_norm": 155.2710418701172,
|
| 196 |
+
"learning_rate": 9.060931899641577e-06,
|
| 197 |
+
"loss": 1.7155,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.0,
|
| 202 |
+
"grad_norm": 165.6481170654297,
|
| 203 |
+
"learning_rate": 9.025089605734768e-06,
|
| 204 |
+
"loss": 1.6922,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 1.04,
|
| 209 |
+
"grad_norm": 108.59546661376953,
|
| 210 |
+
"learning_rate": 8.989247311827958e-06,
|
| 211 |
+
"loss": 1.737,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.08,
|
| 216 |
+
"grad_norm": 119.43255615234375,
|
| 217 |
+
"learning_rate": 8.953405017921147e-06,
|
| 218 |
+
"loss": 1.5624,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.11,
|
| 223 |
+
"grad_norm": 203.9807891845703,
|
| 224 |
+
"learning_rate": 8.917562724014338e-06,
|
| 225 |
+
"loss": 1.5555,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.15,
|
| 230 |
+
"grad_norm": 196.70803833007812,
|
| 231 |
+
"learning_rate": 8.881720430107528e-06,
|
| 232 |
+
"loss": 1.4945,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.18,
|
| 237 |
+
"grad_norm": 218.3573760986328,
|
| 238 |
+
"learning_rate": 8.845878136200717e-06,
|
| 239 |
+
"loss": 1.6784,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.22,
|
| 244 |
+
"grad_norm": 218.7974395751953,
|
| 245 |
+
"learning_rate": 8.810035842293907e-06,
|
| 246 |
+
"loss": 1.5671,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.25,
|
| 251 |
+
"grad_norm": 110.03742980957031,
|
| 252 |
+
"learning_rate": 8.774193548387098e-06,
|
| 253 |
+
"loss": 1.5509,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.29,
|
| 258 |
+
"grad_norm": 536.1112060546875,
|
| 259 |
+
"learning_rate": 8.738351254480287e-06,
|
| 260 |
+
"loss": 1.5836,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.33,
|
| 265 |
+
"grad_norm": 91.13652038574219,
|
| 266 |
+
"learning_rate": 8.702508960573477e-06,
|
| 267 |
+
"loss": 1.5235,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.36,
|
| 272 |
+
"grad_norm": 111.45140075683594,
|
| 273 |
+
"learning_rate": 8.666666666666668e-06,
|
| 274 |
+
"loss": 1.5352,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.4,
|
| 279 |
+
"grad_norm": 83.86921691894531,
|
| 280 |
+
"learning_rate": 8.630824372759857e-06,
|
| 281 |
+
"loss": 1.5994,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.43,
|
| 286 |
+
"grad_norm": 64.33961486816406,
|
| 287 |
+
"learning_rate": 8.594982078853047e-06,
|
| 288 |
+
"loss": 1.5309,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.47,
|
| 293 |
+
"grad_norm": 74.24256134033203,
|
| 294 |
+
"learning_rate": 8.559139784946238e-06,
|
| 295 |
+
"loss": 1.5243,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.51,
|
| 300 |
+
"grad_norm": 88.8588638305664,
|
| 301 |
+
"learning_rate": 8.523297491039427e-06,
|
| 302 |
+
"loss": 1.4779,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.54,
|
| 307 |
+
"grad_norm": 99.83770751953125,
|
| 308 |
+
"learning_rate": 8.487455197132617e-06,
|
| 309 |
+
"loss": 1.4671,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.58,
|
| 314 |
+
"grad_norm": 219.4841766357422,
|
| 315 |
+
"learning_rate": 8.451612903225808e-06,
|
| 316 |
+
"loss": 1.4821,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.61,
|
| 321 |
+
"grad_norm": 145.25010681152344,
|
| 322 |
+
"learning_rate": 8.415770609318998e-06,
|
| 323 |
+
"loss": 1.5258,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.65,
|
| 328 |
+
"grad_norm": 109.16207885742188,
|
| 329 |
+
"learning_rate": 8.379928315412187e-06,
|
| 330 |
+
"loss": 1.4973,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.68,
|
| 335 |
+
"grad_norm": 385.5920104980469,
|
| 336 |
+
"learning_rate": 8.344086021505376e-06,
|
| 337 |
+
"loss": 1.5602,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.72,
|
| 342 |
+
"grad_norm": 78.8775634765625,
|
| 343 |
+
"learning_rate": 8.308243727598568e-06,
|
| 344 |
+
"loss": 1.4304,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.76,
|
| 349 |
+
"grad_norm": 54.696563720703125,
|
| 350 |
+
"learning_rate": 8.272401433691757e-06,
|
| 351 |
+
"loss": 1.4581,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.79,
|
| 356 |
+
"grad_norm": 145.9872283935547,
|
| 357 |
+
"learning_rate": 8.236559139784947e-06,
|
| 358 |
+
"loss": 1.4707,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.83,
|
| 363 |
+
"grad_norm": 86.8324203491211,
|
| 364 |
+
"learning_rate": 8.200716845878138e-06,
|
| 365 |
+
"loss": 1.4593,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.86,
|
| 370 |
+
"grad_norm": 328.5557556152344,
|
| 371 |
+
"learning_rate": 8.164874551971327e-06,
|
| 372 |
+
"loss": 1.4688,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.9,
|
| 377 |
+
"grad_norm": 126.05339813232422,
|
| 378 |
+
"learning_rate": 8.129032258064517e-06,
|
| 379 |
+
"loss": 1.5903,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.94,
|
| 384 |
+
"grad_norm": 216.65573120117188,
|
| 385 |
+
"learning_rate": 8.093189964157708e-06,
|
| 386 |
+
"loss": 1.6609,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.97,
|
| 391 |
+
"grad_norm": 150.45306396484375,
|
| 392 |
+
"learning_rate": 8.057347670250897e-06,
|
| 393 |
+
"loss": 1.5142,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 2.01,
|
| 398 |
+
"grad_norm": 84.9225845336914,
|
| 399 |
+
"learning_rate": 8.021505376344087e-06,
|
| 400 |
+
"loss": 1.5273,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 2.04,
|
| 405 |
+
"grad_norm": 63.36981201171875,
|
| 406 |
+
"learning_rate": 7.985663082437278e-06,
|
| 407 |
+
"loss": 1.4743,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 2.08,
|
| 412 |
+
"grad_norm": 70.1873779296875,
|
| 413 |
+
"learning_rate": 7.949820788530466e-06,
|
| 414 |
+
"loss": 1.5223,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 2.11,
|
| 419 |
+
"grad_norm": 208.42889404296875,
|
| 420 |
+
"learning_rate": 7.913978494623657e-06,
|
| 421 |
+
"loss": 1.4587,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 2.15,
|
| 426 |
+
"grad_norm": 59.578243255615234,
|
| 427 |
+
"learning_rate": 7.878136200716846e-06,
|
| 428 |
+
"loss": 1.3852,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 2.19,
|
| 433 |
+
"grad_norm": 121.19114685058594,
|
| 434 |
+
"learning_rate": 7.842293906810036e-06,
|
| 435 |
+
"loss": 1.3795,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 2.22,
|
| 440 |
+
"grad_norm": 81.8535385131836,
|
| 441 |
+
"learning_rate": 7.806451612903227e-06,
|
| 442 |
+
"loss": 1.3634,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 2.26,
|
| 447 |
+
"grad_norm": 171.9092559814453,
|
| 448 |
+
"learning_rate": 7.770609318996416e-06,
|
| 449 |
+
"loss": 1.4038,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.29,
|
| 454 |
+
"grad_norm": 189.8422393798828,
|
| 455 |
+
"learning_rate": 7.734767025089606e-06,
|
| 456 |
+
"loss": 1.359,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.33,
|
| 461 |
+
"grad_norm": 191.68310546875,
|
| 462 |
+
"learning_rate": 7.698924731182797e-06,
|
| 463 |
+
"loss": 1.4473,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.37,
|
| 468 |
+
"grad_norm": 60.011138916015625,
|
| 469 |
+
"learning_rate": 7.663082437275985e-06,
|
| 470 |
+
"loss": 1.3605,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.4,
|
| 475 |
+
"grad_norm": 85.5229721069336,
|
| 476 |
+
"learning_rate": 7.627240143369177e-06,
|
| 477 |
+
"loss": 1.3775,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.44,
|
| 482 |
+
"grad_norm": 72.5903549194336,
|
| 483 |
+
"learning_rate": 7.5913978494623665e-06,
|
| 484 |
+
"loss": 1.3748,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.47,
|
| 489 |
+
"grad_norm": 246.1010284423828,
|
| 490 |
+
"learning_rate": 7.555555555555556e-06,
|
| 491 |
+
"loss": 1.3974,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.51,
|
| 496 |
+
"grad_norm": 70.74227905273438,
|
| 497 |
+
"learning_rate": 7.519713261648746e-06,
|
| 498 |
+
"loss": 1.352,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.54,
|
| 503 |
+
"grad_norm": 131.19888305664062,
|
| 504 |
+
"learning_rate": 7.483870967741936e-06,
|
| 505 |
+
"loss": 1.3262,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.58,
|
| 510 |
+
"grad_norm": 178.96922302246094,
|
| 511 |
+
"learning_rate": 7.448028673835126e-06,
|
| 512 |
+
"loss": 1.3966,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.62,
|
| 517 |
+
"grad_norm": 173.15139770507812,
|
| 518 |
+
"learning_rate": 7.412186379928316e-06,
|
| 519 |
+
"loss": 1.3522,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.65,
|
| 524 |
+
"grad_norm": 139.38653564453125,
|
| 525 |
+
"learning_rate": 7.376344086021506e-06,
|
| 526 |
+
"loss": 1.3824,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.69,
|
| 531 |
+
"grad_norm": 91.99787902832031,
|
| 532 |
+
"learning_rate": 7.340501792114696e-06,
|
| 533 |
+
"loss": 1.3789,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.72,
|
| 538 |
+
"grad_norm": 113.1898193359375,
|
| 539 |
+
"learning_rate": 7.3046594982078856e-06,
|
| 540 |
+
"loss": 1.4338,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.76,
|
| 545 |
+
"grad_norm": 92.90975952148438,
|
| 546 |
+
"learning_rate": 7.268817204301076e-06,
|
| 547 |
+
"loss": 1.3135,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.8,
|
| 552 |
+
"grad_norm": 62.657371520996094,
|
| 553 |
+
"learning_rate": 7.232974910394266e-06,
|
| 554 |
+
"loss": 1.3329,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.83,
|
| 559 |
+
"grad_norm": 44.4366340637207,
|
| 560 |
+
"learning_rate": 7.1971326164874554e-06,
|
| 561 |
+
"loss": 1.372,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.87,
|
| 566 |
+
"grad_norm": 73.2843017578125,
|
| 567 |
+
"learning_rate": 7.161290322580646e-06,
|
| 568 |
+
"loss": 1.4003,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.9,
|
| 573 |
+
"grad_norm": 33.83133316040039,
|
| 574 |
+
"learning_rate": 7.125448028673836e-06,
|
| 575 |
+
"loss": 1.3719,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.94,
|
| 580 |
+
"grad_norm": 95.57901763916016,
|
| 581 |
+
"learning_rate": 7.089605734767025e-06,
|
| 582 |
+
"loss": 1.3175,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.97,
|
| 587 |
+
"grad_norm": 90.033935546875,
|
| 588 |
+
"learning_rate": 7.053763440860215e-06,
|
| 589 |
+
"loss": 1.3899,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 3.01,
|
| 594 |
+
"grad_norm": 73.42399597167969,
|
| 595 |
+
"learning_rate": 7.0179211469534055e-06,
|
| 596 |
+
"loss": 1.3479,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 3.05,
|
| 601 |
+
"grad_norm": 90.77163696289062,
|
| 602 |
+
"learning_rate": 6.982078853046595e-06,
|
| 603 |
+
"loss": 1.3747,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 3.08,
|
| 608 |
+
"grad_norm": 73.06351470947266,
|
| 609 |
+
"learning_rate": 6.946236559139785e-06,
|
| 610 |
+
"loss": 1.3437,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 3.12,
|
| 615 |
+
"grad_norm": 177.39906311035156,
|
| 616 |
+
"learning_rate": 6.910394265232976e-06,
|
| 617 |
+
"loss": 1.3008,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 3.15,
|
| 622 |
+
"grad_norm": 49.48398208618164,
|
| 623 |
+
"learning_rate": 6.874551971326166e-06,
|
| 624 |
+
"loss": 1.3858,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 3.19,
|
| 629 |
+
"grad_norm": 60.6556396484375,
|
| 630 |
+
"learning_rate": 6.838709677419355e-06,
|
| 631 |
+
"loss": 1.3837,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 3.23,
|
| 636 |
+
"grad_norm": 133.91407775878906,
|
| 637 |
+
"learning_rate": 6.802867383512546e-06,
|
| 638 |
+
"loss": 1.3502,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 3.26,
|
| 643 |
+
"grad_norm": 100.1251449584961,
|
| 644 |
+
"learning_rate": 6.767025089605736e-06,
|
| 645 |
+
"loss": 1.3186,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 3.3,
|
| 650 |
+
"grad_norm": 56.571807861328125,
|
| 651 |
+
"learning_rate": 6.731182795698925e-06,
|
| 652 |
+
"loss": 1.2846,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 3.33,
|
| 657 |
+
"grad_norm": 37.74870681762695,
|
| 658 |
+
"learning_rate": 6.695340501792115e-06,
|
| 659 |
+
"loss": 1.3624,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 3.37,
|
| 664 |
+
"grad_norm": 334.59722900390625,
|
| 665 |
+
"learning_rate": 6.659498207885306e-06,
|
| 666 |
+
"loss": 1.3377,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 3.41,
|
| 671 |
+
"grad_norm": 67.167724609375,
|
| 672 |
+
"learning_rate": 6.623655913978495e-06,
|
| 673 |
+
"loss": 1.3796,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.44,
|
| 678 |
+
"grad_norm": 46.475154876708984,
|
| 679 |
+
"learning_rate": 6.587813620071685e-06,
|
| 680 |
+
"loss": 1.3168,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.48,
|
| 685 |
+
"grad_norm": 94.3563232421875,
|
| 686 |
+
"learning_rate": 6.5519713261648755e-06,
|
| 687 |
+
"loss": 1.2965,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.51,
|
| 692 |
+
"grad_norm": 49.33637619018555,
|
| 693 |
+
"learning_rate": 6.516129032258065e-06,
|
| 694 |
+
"loss": 1.2851,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.55,
|
| 699 |
+
"grad_norm": 78.11804962158203,
|
| 700 |
+
"learning_rate": 6.480286738351255e-06,
|
| 701 |
+
"loss": 1.4281,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.58,
|
| 706 |
+
"grad_norm": 135.7293243408203,
|
| 707 |
+
"learning_rate": 6.444444444444445e-06,
|
| 708 |
+
"loss": 1.361,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.62,
|
| 713 |
+
"grad_norm": 449.07757568359375,
|
| 714 |
+
"learning_rate": 6.408602150537635e-06,
|
| 715 |
+
"loss": 1.3035,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.66,
|
| 720 |
+
"grad_norm": 50.195037841796875,
|
| 721 |
+
"learning_rate": 6.372759856630825e-06,
|
| 722 |
+
"loss": 1.3036,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.69,
|
| 727 |
+
"grad_norm": 51.2278938293457,
|
| 728 |
+
"learning_rate": 6.336917562724015e-06,
|
| 729 |
+
"loss": 1.3152,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.73,
|
| 734 |
+
"grad_norm": 115.5573501586914,
|
| 735 |
+
"learning_rate": 6.301075268817205e-06,
|
| 736 |
+
"loss": 1.2863,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.76,
|
| 741 |
+
"grad_norm": 48.20037078857422,
|
| 742 |
+
"learning_rate": 6.2652329749103945e-06,
|
| 743 |
+
"loss": 1.289,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.8,
|
| 748 |
+
"grad_norm": 82.9049301147461,
|
| 749 |
+
"learning_rate": 6.229390681003584e-06,
|
| 750 |
+
"loss": 1.3162,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.84,
|
| 755 |
+
"grad_norm": 49.85783386230469,
|
| 756 |
+
"learning_rate": 6.193548387096775e-06,
|
| 757 |
+
"loss": 1.439,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.87,
|
| 762 |
+
"grad_norm": 72.38436126708984,
|
| 763 |
+
"learning_rate": 6.157706093189964e-06,
|
| 764 |
+
"loss": 1.3553,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.91,
|
| 769 |
+
"grad_norm": 69.86207580566406,
|
| 770 |
+
"learning_rate": 6.121863799283154e-06,
|
| 771 |
+
"loss": 1.3005,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.94,
|
| 776 |
+
"grad_norm": 121.75460052490234,
|
| 777 |
+
"learning_rate": 6.086021505376345e-06,
|
| 778 |
+
"loss": 1.3552,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.98,
|
| 783 |
+
"grad_norm": 74.01500701904297,
|
| 784 |
+
"learning_rate": 6.050179211469534e-06,
|
| 785 |
+
"loss": 1.2324,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 4.01,
|
| 790 |
+
"grad_norm": 120.46449279785156,
|
| 791 |
+
"learning_rate": 6.014336917562724e-06,
|
| 792 |
+
"loss": 1.2653,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 4.05,
|
| 797 |
+
"grad_norm": 129.35137939453125,
|
| 798 |
+
"learning_rate": 5.978494623655915e-06,
|
| 799 |
+
"loss": 1.2437,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 4.09,
|
| 804 |
+
"grad_norm": 114.84891510009766,
|
| 805 |
+
"learning_rate": 5.942652329749104e-06,
|
| 806 |
+
"loss": 1.2631,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 4.12,
|
| 811 |
+
"grad_norm": 48.180904388427734,
|
| 812 |
+
"learning_rate": 5.906810035842294e-06,
|
| 813 |
+
"loss": 1.2735,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 4.16,
|
| 818 |
+
"grad_norm": 40.2717170715332,
|
| 819 |
+
"learning_rate": 5.8709677419354835e-06,
|
| 820 |
+
"loss": 1.2728,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 4.19,
|
| 825 |
+
"grad_norm": 71.15570068359375,
|
| 826 |
+
"learning_rate": 5.835125448028675e-06,
|
| 827 |
+
"loss": 1.3059,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 4.23,
|
| 832 |
+
"grad_norm": 62.20613479614258,
|
| 833 |
+
"learning_rate": 5.7992831541218645e-06,
|
| 834 |
+
"loss": 1.2574,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 4.27,
|
| 839 |
+
"grad_norm": 83.15790557861328,
|
| 840 |
+
"learning_rate": 5.763440860215054e-06,
|
| 841 |
+
"loss": 1.3026,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 4.3,
|
| 846 |
+
"grad_norm": 79.50647735595703,
|
| 847 |
+
"learning_rate": 5.727598566308245e-06,
|
| 848 |
+
"loss": 1.237,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 4.34,
|
| 853 |
+
"grad_norm": 69.64093780517578,
|
| 854 |
+
"learning_rate": 5.691756272401434e-06,
|
| 855 |
+
"loss": 1.2479,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 4.37,
|
| 860 |
+
"grad_norm": 40.61994171142578,
|
| 861 |
+
"learning_rate": 5.655913978494624e-06,
|
| 862 |
+
"loss": 1.2842,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 4.41,
|
| 867 |
+
"grad_norm": 53.86878967285156,
|
| 868 |
+
"learning_rate": 5.620071684587815e-06,
|
| 869 |
+
"loss": 1.267,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 4.44,
|
| 874 |
+
"grad_norm": 53.299129486083984,
|
| 875 |
+
"learning_rate": 5.584229390681004e-06,
|
| 876 |
+
"loss": 1.2256,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 4.48,
|
| 881 |
+
"grad_norm": 124.71385192871094,
|
| 882 |
+
"learning_rate": 5.548387096774194e-06,
|
| 883 |
+
"loss": 1.245,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.52,
|
| 888 |
+
"grad_norm": 48.88494873046875,
|
| 889 |
+
"learning_rate": 5.5125448028673844e-06,
|
| 890 |
+
"loss": 1.2428,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.55,
|
| 895 |
+
"grad_norm": 34.74821472167969,
|
| 896 |
+
"learning_rate": 5.476702508960574e-06,
|
| 897 |
+
"loss": 1.241,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 4.59,
|
| 902 |
+
"grad_norm": 92.53400421142578,
|
| 903 |
+
"learning_rate": 5.440860215053764e-06,
|
| 904 |
+
"loss": 1.2655,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.62,
|
| 909 |
+
"grad_norm": 79.42727661132812,
|
| 910 |
+
"learning_rate": 5.4050179211469535e-06,
|
| 911 |
+
"loss": 1.2246,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.66,
|
| 916 |
+
"grad_norm": 42.84294891357422,
|
| 917 |
+
"learning_rate": 5.369175627240144e-06,
|
| 918 |
+
"loss": 1.234,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.7,
|
| 923 |
+
"grad_norm": 43.57499313354492,
|
| 924 |
+
"learning_rate": 5.333333333333334e-06,
|
| 925 |
+
"loss": 1.2637,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.73,
|
| 930 |
+
"grad_norm": 141.61326599121094,
|
| 931 |
+
"learning_rate": 5.297491039426523e-06,
|
| 932 |
+
"loss": 1.2883,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.77,
|
| 937 |
+
"grad_norm": 71.41532897949219,
|
| 938 |
+
"learning_rate": 5.261648745519714e-06,
|
| 939 |
+
"loss": 1.2633,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.8,
|
| 944 |
+
"grad_norm": 56.57035827636719,
|
| 945 |
+
"learning_rate": 5.2258064516129035e-06,
|
| 946 |
+
"loss": 1.238,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.84,
|
| 951 |
+
"grad_norm": 38.39865493774414,
|
| 952 |
+
"learning_rate": 5.189964157706093e-06,
|
| 953 |
+
"loss": 1.2587,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.87,
|
| 958 |
+
"grad_norm": 45.26354217529297,
|
| 959 |
+
"learning_rate": 5.154121863799284e-06,
|
| 960 |
+
"loss": 1.2644,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.91,
|
| 965 |
+
"grad_norm": 62.85947799682617,
|
| 966 |
+
"learning_rate": 5.118279569892473e-06,
|
| 967 |
+
"loss": 1.3197,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.95,
|
| 972 |
+
"grad_norm": 171.12010192871094,
|
| 973 |
+
"learning_rate": 5.082437275985663e-06,
|
| 974 |
+
"loss": 1.2333,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.98,
|
| 979 |
+
"grad_norm": 38.02899932861328,
|
| 980 |
+
"learning_rate": 5.0465949820788544e-06,
|
| 981 |
+
"loss": 1.2987,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 5.02,
|
| 986 |
+
"grad_norm": 47.150367736816406,
|
| 987 |
+
"learning_rate": 5.010752688172043e-06,
|
| 988 |
+
"loss": 1.2651,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 5.05,
|
| 993 |
+
"grad_norm": 50.13650894165039,
|
| 994 |
+
"learning_rate": 4.974910394265233e-06,
|
| 995 |
+
"loss": 1.224,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 5.09,
|
| 1000 |
+
"grad_norm": 56.77107238769531,
|
| 1001 |
+
"learning_rate": 4.9390681003584234e-06,
|
| 1002 |
+
"loss": 1.1944,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 5.13,
|
| 1007 |
+
"grad_norm": 45.99001693725586,
|
| 1008 |
+
"learning_rate": 4.903225806451613e-06,
|
| 1009 |
+
"loss": 1.2481,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 5.16,
|
| 1014 |
+
"grad_norm": 94.43379974365234,
|
| 1015 |
+
"learning_rate": 4.867383512544804e-06,
|
| 1016 |
+
"loss": 1.2333,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 5.2,
|
| 1021 |
+
"grad_norm": 92.9940414428711,
|
| 1022 |
+
"learning_rate": 4.831541218637993e-06,
|
| 1023 |
+
"loss": 1.2503,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 5.23,
|
| 1028 |
+
"grad_norm": 62.211280822753906,
|
| 1029 |
+
"learning_rate": 4.795698924731183e-06,
|
| 1030 |
+
"loss": 1.238,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 5.27,
|
| 1035 |
+
"grad_norm": 56.011722564697266,
|
| 1036 |
+
"learning_rate": 4.7598566308243735e-06,
|
| 1037 |
+
"loss": 1.2228,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 5.3,
|
| 1042 |
+
"grad_norm": 131.8081817626953,
|
| 1043 |
+
"learning_rate": 4.724014336917563e-06,
|
| 1044 |
+
"loss": 1.1854,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 5.34,
|
| 1049 |
+
"grad_norm": 38.84933090209961,
|
| 1050 |
+
"learning_rate": 4.688172043010753e-06,
|
| 1051 |
+
"loss": 1.2052,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 5.38,
|
| 1056 |
+
"grad_norm": 53.045658111572266,
|
| 1057 |
+
"learning_rate": 4.652329749103943e-06,
|
| 1058 |
+
"loss": 1.3113,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 5.41,
|
| 1063 |
+
"grad_norm": 48.19401931762695,
|
| 1064 |
+
"learning_rate": 4.616487455197133e-06,
|
| 1065 |
+
"loss": 1.2192,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 5.45,
|
| 1070 |
+
"grad_norm": 31.12712860107422,
|
| 1071 |
+
"learning_rate": 4.580645161290323e-06,
|
| 1072 |
+
"loss": 1.2003,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 5.48,
|
| 1077 |
+
"grad_norm": 2096.69091796875,
|
| 1078 |
+
"learning_rate": 4.544802867383513e-06,
|
| 1079 |
+
"loss": 1.2451,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 5.52,
|
| 1084 |
+
"grad_norm": 76.25735473632812,
|
| 1085 |
+
"learning_rate": 4.508960573476703e-06,
|
| 1086 |
+
"loss": 1.2152,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 5.56,
|
| 1091 |
+
"grad_norm": 66.96879577636719,
|
| 1092 |
+
"learning_rate": 4.473118279569893e-06,
|
| 1093 |
+
"loss": 1.1985,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 5.59,
|
| 1098 |
+
"grad_norm": 74.71087646484375,
|
| 1099 |
+
"learning_rate": 4.437275985663082e-06,
|
| 1100 |
+
"loss": 1.2033,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 5.63,
|
| 1105 |
+
"grad_norm": 64.05677795410156,
|
| 1106 |
+
"learning_rate": 4.401433691756273e-06,
|
| 1107 |
+
"loss": 1.2461,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 5.66,
|
| 1112 |
+
"grad_norm": 35.51209259033203,
|
| 1113 |
+
"learning_rate": 4.365591397849463e-06,
|
| 1114 |
+
"loss": 1.2277,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 5.7,
|
| 1119 |
+
"grad_norm": 50.55402374267578,
|
| 1120 |
+
"learning_rate": 4.329749103942653e-06,
|
| 1121 |
+
"loss": 1.2312,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 5.73,
|
| 1126 |
+
"grad_norm": 39.16292953491211,
|
| 1127 |
+
"learning_rate": 4.293906810035843e-06,
|
| 1128 |
+
"loss": 1.1947,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 5.77,
|
| 1133 |
+
"grad_norm": 108.2676773071289,
|
| 1134 |
+
"learning_rate": 4.258064516129032e-06,
|
| 1135 |
+
"loss": 1.2399,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 5.81,
|
| 1140 |
+
"grad_norm": 37.47825622558594,
|
| 1141 |
+
"learning_rate": 4.222222222222223e-06,
|
| 1142 |
+
"loss": 1.2542,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 5.84,
|
| 1147 |
+
"grad_norm": 41.715301513671875,
|
| 1148 |
+
"learning_rate": 4.1863799283154125e-06,
|
| 1149 |
+
"loss": 1.1657,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.88,
|
| 1154 |
+
"grad_norm": 70.7906723022461,
|
| 1155 |
+
"learning_rate": 4.150537634408602e-06,
|
| 1156 |
+
"loss": 1.18,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.91,
|
| 1161 |
+
"grad_norm": 73.00814056396484,
|
| 1162 |
+
"learning_rate": 4.114695340501793e-06,
|
| 1163 |
+
"loss": 1.1753,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.95,
|
| 1168 |
+
"grad_norm": 40.9326286315918,
|
| 1169 |
+
"learning_rate": 4.078853046594982e-06,
|
| 1170 |
+
"loss": 1.1954,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.99,
|
| 1175 |
+
"grad_norm": 71.95938873291016,
|
| 1176 |
+
"learning_rate": 4.043010752688172e-06,
|
| 1177 |
+
"loss": 1.239,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 6.02,
|
| 1182 |
+
"grad_norm": 35.70452880859375,
|
| 1183 |
+
"learning_rate": 4.0071684587813626e-06,
|
| 1184 |
+
"loss": 1.2607,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 6.06,
|
| 1189 |
+
"grad_norm": 33.75741958618164,
|
| 1190 |
+
"learning_rate": 3.971326164874552e-06,
|
| 1191 |
+
"loss": 1.2169,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 6.09,
|
| 1196 |
+
"grad_norm": 39.85159683227539,
|
| 1197 |
+
"learning_rate": 3.935483870967742e-06,
|
| 1198 |
+
"loss": 1.198,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 6.13,
|
| 1203 |
+
"grad_norm": 51.76079177856445,
|
| 1204 |
+
"learning_rate": 3.8996415770609324e-06,
|
| 1205 |
+
"loss": 1.2028,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 6.16,
|
| 1210 |
+
"grad_norm": 41.39125061035156,
|
| 1211 |
+
"learning_rate": 3.863799283154122e-06,
|
| 1212 |
+
"loss": 1.2141,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 6.2,
|
| 1217 |
+
"grad_norm": 53.27019119262695,
|
| 1218 |
+
"learning_rate": 3.827956989247313e-06,
|
| 1219 |
+
"loss": 1.1815,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 6.24,
|
| 1224 |
+
"grad_norm": 54.405303955078125,
|
| 1225 |
+
"learning_rate": 3.792114695340502e-06,
|
| 1226 |
+
"loss": 1.2037,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 6.27,
|
| 1231 |
+
"grad_norm": 60.07841491699219,
|
| 1232 |
+
"learning_rate": 3.756272401433692e-06,
|
| 1233 |
+
"loss": 1.2098,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 6.31,
|
| 1238 |
+
"grad_norm": 82.04875183105469,
|
| 1239 |
+
"learning_rate": 3.720430107526882e-06,
|
| 1240 |
+
"loss": 1.1644,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 6.34,
|
| 1245 |
+
"grad_norm": 26.22403335571289,
|
| 1246 |
+
"learning_rate": 3.6845878136200717e-06,
|
| 1247 |
+
"loss": 1.1614,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 6.38,
|
| 1252 |
+
"grad_norm": 48.65098190307617,
|
| 1253 |
+
"learning_rate": 3.6487455197132623e-06,
|
| 1254 |
+
"loss": 1.2599,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 6.42,
|
| 1259 |
+
"grad_norm": 42.07708740234375,
|
| 1260 |
+
"learning_rate": 3.6129032258064515e-06,
|
| 1261 |
+
"loss": 1.0894,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 6.45,
|
| 1266 |
+
"grad_norm": 36.60328674316406,
|
| 1267 |
+
"learning_rate": 3.577060931899642e-06,
|
| 1268 |
+
"loss": 1.1918,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 6.49,
|
| 1273 |
+
"grad_norm": 61.540828704833984,
|
| 1274 |
+
"learning_rate": 3.541218637992832e-06,
|
| 1275 |
+
"loss": 1.1573,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 6.52,
|
| 1280 |
+
"grad_norm": 55.06793975830078,
|
| 1281 |
+
"learning_rate": 3.505376344086022e-06,
|
| 1282 |
+
"loss": 1.2161,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 6.56,
|
| 1287 |
+
"grad_norm": 52.780025482177734,
|
| 1288 |
+
"learning_rate": 3.469534050179212e-06,
|
| 1289 |
+
"loss": 1.2136,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 6.59,
|
| 1294 |
+
"grad_norm": 36.805023193359375,
|
| 1295 |
+
"learning_rate": 3.4336917562724016e-06,
|
| 1296 |
+
"loss": 1.2059,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 6.63,
|
| 1301 |
+
"grad_norm": 34.84747314453125,
|
| 1302 |
+
"learning_rate": 3.3978494623655917e-06,
|
| 1303 |
+
"loss": 1.1723,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 6.67,
|
| 1308 |
+
"grad_norm": 29.03413200378418,
|
| 1309 |
+
"learning_rate": 3.3620071684587818e-06,
|
| 1310 |
+
"loss": 1.2304,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 6.7,
|
| 1315 |
+
"grad_norm": 43.57373809814453,
|
| 1316 |
+
"learning_rate": 3.3261648745519714e-06,
|
| 1317 |
+
"loss": 1.2636,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 6.74,
|
| 1322 |
+
"grad_norm": 40.338401794433594,
|
| 1323 |
+
"learning_rate": 3.2903225806451615e-06,
|
| 1324 |
+
"loss": 1.201,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 6.77,
|
| 1329 |
+
"grad_norm": 49.39339828491211,
|
| 1330 |
+
"learning_rate": 3.254480286738351e-06,
|
| 1331 |
+
"loss": 1.1907,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 6.81,
|
| 1336 |
+
"grad_norm": 44.37787628173828,
|
| 1337 |
+
"learning_rate": 3.2186379928315413e-06,
|
| 1338 |
+
"loss": 1.1608,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 6.85,
|
| 1343 |
+
"grad_norm": 30.73851776123047,
|
| 1344 |
+
"learning_rate": 3.182795698924732e-06,
|
| 1345 |
+
"loss": 1.0991,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 6.88,
|
| 1350 |
+
"grad_norm": 75.66316986083984,
|
| 1351 |
+
"learning_rate": 3.146953405017921e-06,
|
| 1352 |
+
"loss": 1.2072,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 6.92,
|
| 1357 |
+
"grad_norm": 79.0891342163086,
|
| 1358 |
+
"learning_rate": 3.1111111111111116e-06,
|
| 1359 |
+
"loss": 1.1899,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 6.95,
|
| 1364 |
+
"grad_norm": 44.473907470703125,
|
| 1365 |
+
"learning_rate": 3.0752688172043017e-06,
|
| 1366 |
+
"loss": 1.2603,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 6.99,
|
| 1371 |
+
"grad_norm": 70.97260284423828,
|
| 1372 |
+
"learning_rate": 3.0394265232974914e-06,
|
| 1373 |
+
"loss": 1.1917,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 7.03,
|
| 1378 |
+
"grad_norm": 113.41941833496094,
|
| 1379 |
+
"learning_rate": 3.0035842293906814e-06,
|
| 1380 |
+
"loss": 1.2268,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 7.06,
|
| 1385 |
+
"grad_norm": 31.316911697387695,
|
| 1386 |
+
"learning_rate": 2.967741935483871e-06,
|
| 1387 |
+
"loss": 1.1049,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 7.1,
|
| 1392 |
+
"grad_norm": 35.60115051269531,
|
| 1393 |
+
"learning_rate": 2.9318996415770612e-06,
|
| 1394 |
+
"loss": 1.1542,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 7.13,
|
| 1399 |
+
"grad_norm": 57.42852783203125,
|
| 1400 |
+
"learning_rate": 2.8960573476702513e-06,
|
| 1401 |
+
"loss": 1.2427,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 7.17,
|
| 1406 |
+
"grad_norm": 40.23640441894531,
|
| 1407 |
+
"learning_rate": 2.860215053763441e-06,
|
| 1408 |
+
"loss": 1.1448,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 7.2,
|
| 1413 |
+
"grad_norm": 36.7147102355957,
|
| 1414 |
+
"learning_rate": 2.824372759856631e-06,
|
| 1415 |
+
"loss": 1.1392,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 7.24,
|
| 1420 |
+
"grad_norm": 33.4290657043457,
|
| 1421 |
+
"learning_rate": 2.7885304659498208e-06,
|
| 1422 |
+
"loss": 1.1704,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 7.28,
|
| 1427 |
+
"grad_norm": 56.32290267944336,
|
| 1428 |
+
"learning_rate": 2.752688172043011e-06,
|
| 1429 |
+
"loss": 1.1495,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 7.31,
|
| 1434 |
+
"grad_norm": 43.89424133300781,
|
| 1435 |
+
"learning_rate": 2.716845878136201e-06,
|
| 1436 |
+
"loss": 1.1387,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 7.35,
|
| 1441 |
+
"grad_norm": 39.107975006103516,
|
| 1442 |
+
"learning_rate": 2.6810035842293906e-06,
|
| 1443 |
+
"loss": 1.3479,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 7.38,
|
| 1448 |
+
"grad_norm": 56.70566177368164,
|
| 1449 |
+
"learning_rate": 2.645161290322581e-06,
|
| 1450 |
+
"loss": 1.1315,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 7.42,
|
| 1455 |
+
"grad_norm": 28.802082061767578,
|
| 1456 |
+
"learning_rate": 2.6093189964157704e-06,
|
| 1457 |
+
"loss": 1.1935,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 7.46,
|
| 1462 |
+
"grad_norm": 48.39360809326172,
|
| 1463 |
+
"learning_rate": 2.573476702508961e-06,
|
| 1464 |
+
"loss": 1.2482,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 7.49,
|
| 1469 |
+
"grad_norm": 32.54924774169922,
|
| 1470 |
+
"learning_rate": 2.537634408602151e-06,
|
| 1471 |
+
"loss": 1.2273,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 7.53,
|
| 1476 |
+
"grad_norm": 39.42136764526367,
|
| 1477 |
+
"learning_rate": 2.5017921146953407e-06,
|
| 1478 |
+
"loss": 1.1796,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 7.56,
|
| 1483 |
+
"grad_norm": 45.770755767822266,
|
| 1484 |
+
"learning_rate": 2.4659498207885308e-06,
|
| 1485 |
+
"loss": 1.1732,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 7.6,
|
| 1490 |
+
"grad_norm": 38.276832580566406,
|
| 1491 |
+
"learning_rate": 2.4301075268817204e-06,
|
| 1492 |
+
"loss": 1.1732,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 7.63,
|
| 1497 |
+
"grad_norm": 41.075775146484375,
|
| 1498 |
+
"learning_rate": 2.3942652329749105e-06,
|
| 1499 |
+
"loss": 1.1713,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 7.67,
|
| 1504 |
+
"grad_norm": 65.88035583496094,
|
| 1505 |
+
"learning_rate": 2.3584229390681006e-06,
|
| 1506 |
+
"loss": 1.1167,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 7.71,
|
| 1511 |
+
"grad_norm": 36.76555633544922,
|
| 1512 |
+
"learning_rate": 2.3225806451612907e-06,
|
| 1513 |
+
"loss": 1.1947,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 7.74,
|
| 1518 |
+
"grad_norm": 60.711055755615234,
|
| 1519 |
+
"learning_rate": 2.2867383512544804e-06,
|
| 1520 |
+
"loss": 1.1164,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 7.78,
|
| 1525 |
+
"grad_norm": 37.08338928222656,
|
| 1526 |
+
"learning_rate": 2.2508960573476705e-06,
|
| 1527 |
+
"loss": 1.1199,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 7.81,
|
| 1532 |
+
"grad_norm": 41.63785171508789,
|
| 1533 |
+
"learning_rate": 2.21505376344086e-06,
|
| 1534 |
+
"loss": 1.2048,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 7.85,
|
| 1539 |
+
"grad_norm": 62.35722351074219,
|
| 1540 |
+
"learning_rate": 2.1792114695340507e-06,
|
| 1541 |
+
"loss": 1.2011,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 7.89,
|
| 1546 |
+
"grad_norm": 44.37682342529297,
|
| 1547 |
+
"learning_rate": 2.1433691756272404e-06,
|
| 1548 |
+
"loss": 1.1709,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 7.92,
|
| 1553 |
+
"grad_norm": 40.216739654541016,
|
| 1554 |
+
"learning_rate": 2.1075268817204305e-06,
|
| 1555 |
+
"loss": 1.1959,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 7.96,
|
| 1560 |
+
"grad_norm": 35.443302154541016,
|
| 1561 |
+
"learning_rate": 2.07168458781362e-06,
|
| 1562 |
+
"loss": 1.1123,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 7.99,
|
| 1567 |
+
"grad_norm": 59.596954345703125,
|
| 1568 |
+
"learning_rate": 2.0358422939068102e-06,
|
| 1569 |
+
"loss": 1.1139,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 8.03,
|
| 1574 |
+
"grad_norm": 34.79387283325195,
|
| 1575 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 1576 |
+
"loss": 1.1845,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 8.06,
|
| 1581 |
+
"grad_norm": 50.571712493896484,
|
| 1582 |
+
"learning_rate": 1.96415770609319e-06,
|
| 1583 |
+
"loss": 1.1528,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 8.1,
|
| 1588 |
+
"grad_norm": 33.95150375366211,
|
| 1589 |
+
"learning_rate": 1.92831541218638e-06,
|
| 1590 |
+
"loss": 1.1858,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 8.14,
|
| 1595 |
+
"grad_norm": 64.71048736572266,
|
| 1596 |
+
"learning_rate": 1.89247311827957e-06,
|
| 1597 |
+
"loss": 1.1199,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 8.17,
|
| 1602 |
+
"grad_norm": 76.19322967529297,
|
| 1603 |
+
"learning_rate": 1.8566308243727599e-06,
|
| 1604 |
+
"loss": 1.1457,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 8.21,
|
| 1609 |
+
"grad_norm": 40.506675720214844,
|
| 1610 |
+
"learning_rate": 1.82078853046595e-06,
|
| 1611 |
+
"loss": 1.1886,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 8.24,
|
| 1616 |
+
"grad_norm": 34.859432220458984,
|
| 1617 |
+
"learning_rate": 1.7849462365591399e-06,
|
| 1618 |
+
"loss": 1.1133,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 8.28,
|
| 1623 |
+
"grad_norm": 39.036376953125,
|
| 1624 |
+
"learning_rate": 1.74910394265233e-06,
|
| 1625 |
+
"loss": 1.1845,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 8.32,
|
| 1630 |
+
"grad_norm": 72.71634674072266,
|
| 1631 |
+
"learning_rate": 1.7132616487455198e-06,
|
| 1632 |
+
"loss": 1.1685,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 8.35,
|
| 1637 |
+
"grad_norm": 29.658227920532227,
|
| 1638 |
+
"learning_rate": 1.67741935483871e-06,
|
| 1639 |
+
"loss": 1.1849,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 8.39,
|
| 1644 |
+
"grad_norm": 40.88108825683594,
|
| 1645 |
+
"learning_rate": 1.6415770609318998e-06,
|
| 1646 |
+
"loss": 1.1473,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 8.42,
|
| 1651 |
+
"grad_norm": 46.79905700683594,
|
| 1652 |
+
"learning_rate": 1.6057347670250897e-06,
|
| 1653 |
+
"loss": 1.1407,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 8.46,
|
| 1658 |
+
"grad_norm": 40.85004806518555,
|
| 1659 |
+
"learning_rate": 1.5698924731182796e-06,
|
| 1660 |
+
"loss": 1.0891,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 8.49,
|
| 1665 |
+
"grad_norm": 43.564849853515625,
|
| 1666 |
+
"learning_rate": 1.5340501792114695e-06,
|
| 1667 |
+
"loss": 1.1173,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 8.53,
|
| 1672 |
+
"grad_norm": 39.988792419433594,
|
| 1673 |
+
"learning_rate": 1.4982078853046598e-06,
|
| 1674 |
+
"loss": 1.0948,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 8.57,
|
| 1679 |
+
"grad_norm": 33.0150260925293,
|
| 1680 |
+
"learning_rate": 1.4623655913978497e-06,
|
| 1681 |
+
"loss": 1.1763,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 8.6,
|
| 1686 |
+
"grad_norm": 36.02336120605469,
|
| 1687 |
+
"learning_rate": 1.4265232974910395e-06,
|
| 1688 |
+
"loss": 1.1621,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 8.64,
|
| 1693 |
+
"grad_norm": 34.898765563964844,
|
| 1694 |
+
"learning_rate": 1.3906810035842294e-06,
|
| 1695 |
+
"loss": 1.2091,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 8.67,
|
| 1700 |
+
"grad_norm": 34.42953109741211,
|
| 1701 |
+
"learning_rate": 1.3548387096774195e-06,
|
| 1702 |
+
"loss": 1.1255,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 8.71,
|
| 1707 |
+
"grad_norm": 86.2882080078125,
|
| 1708 |
+
"learning_rate": 1.3189964157706094e-06,
|
| 1709 |
+
"loss": 1.1501,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 8.75,
|
| 1714 |
+
"grad_norm": 43.11504364013672,
|
| 1715 |
+
"learning_rate": 1.2831541218637993e-06,
|
| 1716 |
+
"loss": 1.2159,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 8.78,
|
| 1721 |
+
"grad_norm": 37.7353630065918,
|
| 1722 |
+
"learning_rate": 1.2473118279569894e-06,
|
| 1723 |
+
"loss": 1.1255,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 8.82,
|
| 1728 |
+
"grad_norm": 33.8388671875,
|
| 1729 |
+
"learning_rate": 1.2114695340501793e-06,
|
| 1730 |
+
"loss": 1.1677,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 8.85,
|
| 1735 |
+
"grad_norm": 129.2806396484375,
|
| 1736 |
+
"learning_rate": 1.1756272401433692e-06,
|
| 1737 |
+
"loss": 1.1739,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 8.89,
|
| 1742 |
+
"grad_norm": 31.244264602661133,
|
| 1743 |
+
"learning_rate": 1.1397849462365593e-06,
|
| 1744 |
+
"loss": 1.1247,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 8.92,
|
| 1749 |
+
"grad_norm": 30.236568450927734,
|
| 1750 |
+
"learning_rate": 1.1039426523297491e-06,
|
| 1751 |
+
"loss": 1.093,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 8.96,
|
| 1756 |
+
"grad_norm": 42.34114456176758,
|
| 1757 |
+
"learning_rate": 1.0681003584229392e-06,
|
| 1758 |
+
"loss": 1.1688,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 9.0,
|
| 1763 |
+
"grad_norm": 34.8842887878418,
|
| 1764 |
+
"learning_rate": 1.0322580645161291e-06,
|
| 1765 |
+
"loss": 1.1828,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 9.03,
|
| 1770 |
+
"grad_norm": 34.398521423339844,
|
| 1771 |
+
"learning_rate": 9.96415770609319e-07,
|
| 1772 |
+
"loss": 1.1351,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 9.07,
|
| 1777 |
+
"grad_norm": 42.38921356201172,
|
| 1778 |
+
"learning_rate": 9.60573476702509e-07,
|
| 1779 |
+
"loss": 1.1673,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 9.1,
|
| 1784 |
+
"grad_norm": 62.44378662109375,
|
| 1785 |
+
"learning_rate": 9.24731182795699e-07,
|
| 1786 |
+
"loss": 1.1386,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 9.14,
|
| 1791 |
+
"grad_norm": 30.45479965209961,
|
| 1792 |
+
"learning_rate": 8.88888888888889e-07,
|
| 1793 |
+
"loss": 1.1577,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 9.18,
|
| 1798 |
+
"grad_norm": 37.57691192626953,
|
| 1799 |
+
"learning_rate": 8.530465949820789e-07,
|
| 1800 |
+
"loss": 1.1291,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 9.21,
|
| 1805 |
+
"grad_norm": 54.45702362060547,
|
| 1806 |
+
"learning_rate": 8.17204301075269e-07,
|
| 1807 |
+
"loss": 1.1575,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 9.25,
|
| 1812 |
+
"grad_norm": 57.186737060546875,
|
| 1813 |
+
"learning_rate": 7.813620071684588e-07,
|
| 1814 |
+
"loss": 1.132,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 9.28,
|
| 1819 |
+
"grad_norm": 35.08234405517578,
|
| 1820 |
+
"learning_rate": 7.455197132616488e-07,
|
| 1821 |
+
"loss": 1.1816,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 9.32,
|
| 1826 |
+
"grad_norm": 87.56676483154297,
|
| 1827 |
+
"learning_rate": 7.096774193548388e-07,
|
| 1828 |
+
"loss": 1.1528,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
}
|
| 1831 |
+
],
|
| 1832 |
+
"logging_steps": 10,
|
| 1833 |
+
"max_steps": 2790,
|
| 1834 |
+
"num_input_tokens_seen": 0,
|
| 1835 |
+
"num_train_epochs": 10,
|
| 1836 |
+
"save_steps": 200,
|
| 1837 |
+
"total_flos": 3.965868410199552e+19,
|
| 1838 |
+
"train_batch_size": 32,
|
| 1839 |
+
"trial_name": null,
|
| 1840 |
+
"trial_params": null
|
| 1841 |
+
}
|