Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2107 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2107 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2950,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.03389830508474576,
|
| 13 |
+
"grad_norm": 4.791459083557129,
|
| 14 |
+
"learning_rate": 6.7567567567567575e-06,
|
| 15 |
+
"loss": 0.2753,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.06779661016949153,
|
| 20 |
+
"grad_norm": 2.2864952087402344,
|
| 21 |
+
"learning_rate": 1.3513513513513515e-05,
|
| 22 |
+
"loss": 0.1531,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.1016949152542373,
|
| 27 |
+
"grad_norm": 0.3792467415332794,
|
| 28 |
+
"learning_rate": 2.0270270270270273e-05,
|
| 29 |
+
"loss": 0.0584,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.13559322033898305,
|
| 34 |
+
"grad_norm": 0.4453708231449127,
|
| 35 |
+
"learning_rate": 2.702702702702703e-05,
|
| 36 |
+
"loss": 0.035,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.1694915254237288,
|
| 41 |
+
"grad_norm": 0.6060521602630615,
|
| 42 |
+
"learning_rate": 3.3783783783783784e-05,
|
| 43 |
+
"loss": 0.0155,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.2033898305084746,
|
| 48 |
+
"grad_norm": 0.4304051995277405,
|
| 49 |
+
"learning_rate": 4.0540540540540545e-05,
|
| 50 |
+
"loss": 0.0089,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.23728813559322035,
|
| 55 |
+
"grad_norm": 0.43952998518943787,
|
| 56 |
+
"learning_rate": 4.72972972972973e-05,
|
| 57 |
+
"loss": 0.0089,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.2711864406779661,
|
| 62 |
+
"grad_norm": 0.3579324781894684,
|
| 63 |
+
"learning_rate": 5.405405405405406e-05,
|
| 64 |
+
"loss": 0.0073,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.3050847457627119,
|
| 69 |
+
"grad_norm": 0.221815824508667,
|
| 70 |
+
"learning_rate": 6.0810810810810814e-05,
|
| 71 |
+
"loss": 0.0061,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.3389830508474576,
|
| 76 |
+
"grad_norm": 0.4653712809085846,
|
| 77 |
+
"learning_rate": 6.756756756756757e-05,
|
| 78 |
+
"loss": 0.0046,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.3728813559322034,
|
| 83 |
+
"grad_norm": 0.20366229116916656,
|
| 84 |
+
"learning_rate": 7.432432432432433e-05,
|
| 85 |
+
"loss": 0.0047,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.4067796610169492,
|
| 90 |
+
"grad_norm": 0.30185437202453613,
|
| 91 |
+
"learning_rate": 8.108108108108109e-05,
|
| 92 |
+
"loss": 0.0048,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.4406779661016949,
|
| 97 |
+
"grad_norm": 0.3719215989112854,
|
| 98 |
+
"learning_rate": 8.783783783783784e-05,
|
| 99 |
+
"loss": 0.0045,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.4745762711864407,
|
| 104 |
+
"grad_norm": 0.14779077470302582,
|
| 105 |
+
"learning_rate": 9.45945945945946e-05,
|
| 106 |
+
"loss": 0.0058,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.5084745762711864,
|
| 111 |
+
"grad_norm": 0.525753915309906,
|
| 112 |
+
"learning_rate": 9.999987429188866e-05,
|
| 113 |
+
"loss": 0.0043,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.5423728813559322,
|
| 118 |
+
"grad_norm": 0.3979560136795044,
|
| 119 |
+
"learning_rate": 9.999547457436221e-05,
|
| 120 |
+
"loss": 0.0051,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.576271186440678,
|
| 125 |
+
"grad_norm": 0.42477351427078247,
|
| 126 |
+
"learning_rate": 9.998479008335589e-05,
|
| 127 |
+
"loss": 0.0037,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.6101694915254238,
|
| 132 |
+
"grad_norm": 0.19697535037994385,
|
| 133 |
+
"learning_rate": 9.996782216198338e-05,
|
| 134 |
+
"loss": 0.0052,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.6440677966101694,
|
| 139 |
+
"grad_norm": 0.23994311690330505,
|
| 140 |
+
"learning_rate": 9.994457294322858e-05,
|
| 141 |
+
"loss": 0.0043,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.6779661016949152,
|
| 146 |
+
"grad_norm": 0.3949716091156006,
|
| 147 |
+
"learning_rate": 9.991504534967746e-05,
|
| 148 |
+
"loss": 0.0043,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.711864406779661,
|
| 153 |
+
"grad_norm": 0.5126088261604309,
|
| 154 |
+
"learning_rate": 9.987924309315075e-05,
|
| 155 |
+
"loss": 0.0029,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.7457627118644068,
|
| 160 |
+
"grad_norm": 0.34605440497398376,
|
| 161 |
+
"learning_rate": 9.983717067423721e-05,
|
| 162 |
+
"loss": 0.0024,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.7796610169491526,
|
| 167 |
+
"grad_norm": 0.3350368142127991,
|
| 168 |
+
"learning_rate": 9.978883338172798e-05,
|
| 169 |
+
"loss": 0.0028,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.8135593220338984,
|
| 174 |
+
"grad_norm": 0.4186588525772095,
|
| 175 |
+
"learning_rate": 9.973423729195168e-05,
|
| 176 |
+
"loss": 0.0037,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.847457627118644,
|
| 181 |
+
"grad_norm": 0.4036902189254761,
|
| 182 |
+
"learning_rate": 9.967338926801067e-05,
|
| 183 |
+
"loss": 0.0031,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.8813559322033898,
|
| 188 |
+
"grad_norm": 0.3429702818393707,
|
| 189 |
+
"learning_rate": 9.960629695891814e-05,
|
| 190 |
+
"loss": 0.0035,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.9152542372881356,
|
| 195 |
+
"grad_norm": 0.22962354123592377,
|
| 196 |
+
"learning_rate": 9.953296879863677e-05,
|
| 197 |
+
"loss": 0.0039,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.9491525423728814,
|
| 202 |
+
"grad_norm": 0.3024398386478424,
|
| 203 |
+
"learning_rate": 9.945341400501838e-05,
|
| 204 |
+
"loss": 0.0025,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.9830508474576272,
|
| 209 |
+
"grad_norm": 0.29498374462127686,
|
| 210 |
+
"learning_rate": 9.936764257864525e-05,
|
| 211 |
+
"loss": 0.0022,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.0169491525423728,
|
| 216 |
+
"grad_norm": 0.2633371651172638,
|
| 217 |
+
"learning_rate": 9.927566530157298e-05,
|
| 218 |
+
"loss": 0.0046,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.0508474576271187,
|
| 223 |
+
"grad_norm": 0.1469630002975464,
|
| 224 |
+
"learning_rate": 9.917749373597504e-05,
|
| 225 |
+
"loss": 0.003,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.0847457627118644,
|
| 230 |
+
"grad_norm": 0.2827948033809662,
|
| 231 |
+
"learning_rate": 9.907314022268946e-05,
|
| 232 |
+
"loss": 0.0025,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.11864406779661,
|
| 237 |
+
"grad_norm": 0.21969269216060638,
|
| 238 |
+
"learning_rate": 9.896261787966739e-05,
|
| 239 |
+
"loss": 0.0029,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.152542372881356,
|
| 244 |
+
"grad_norm": 0.20057785511016846,
|
| 245 |
+
"learning_rate": 9.884594060032406e-05,
|
| 246 |
+
"loss": 0.0021,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.1864406779661016,
|
| 251 |
+
"grad_norm": 0.16977977752685547,
|
| 252 |
+
"learning_rate": 9.872312305179243e-05,
|
| 253 |
+
"loss": 0.0029,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.2203389830508475,
|
| 258 |
+
"grad_norm": 0.25367122888565063,
|
| 259 |
+
"learning_rate": 9.859418067307928e-05,
|
| 260 |
+
"loss": 0.0025,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.2542372881355932,
|
| 265 |
+
"grad_norm": 0.25058019161224365,
|
| 266 |
+
"learning_rate": 9.84591296731245e-05,
|
| 267 |
+
"loss": 0.0024,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.288135593220339,
|
| 272 |
+
"grad_norm": 0.15482617914676666,
|
| 273 |
+
"learning_rate": 9.831798702876352e-05,
|
| 274 |
+
"loss": 0.0027,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.3220338983050848,
|
| 279 |
+
"grad_norm": 0.4145447313785553,
|
| 280 |
+
"learning_rate": 9.817077048259314e-05,
|
| 281 |
+
"loss": 0.0034,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.3559322033898304,
|
| 286 |
+
"grad_norm": 0.21151414513587952,
|
| 287 |
+
"learning_rate": 9.801749854074122e-05,
|
| 288 |
+
"loss": 0.0031,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.3898305084745763,
|
| 293 |
+
"grad_norm": 0.19668647646903992,
|
| 294 |
+
"learning_rate": 9.785819047054035e-05,
|
| 295 |
+
"loss": 0.0015,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.423728813559322,
|
| 300 |
+
"grad_norm": 0.12844844162464142,
|
| 301 |
+
"learning_rate": 9.769286629810572e-05,
|
| 302 |
+
"loss": 0.0021,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.457627118644068,
|
| 307 |
+
"grad_norm": 0.10224123299121857,
|
| 308 |
+
"learning_rate": 9.752154680581783e-05,
|
| 309 |
+
"loss": 0.0017,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.4915254237288136,
|
| 314 |
+
"grad_norm": 0.13156643509864807,
|
| 315 |
+
"learning_rate": 9.73442535297099e-05,
|
| 316 |
+
"loss": 0.0022,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.5254237288135593,
|
| 321 |
+
"grad_norm": 0.09552246332168579,
|
| 322 |
+
"learning_rate": 9.716100875676071e-05,
|
| 323 |
+
"loss": 0.0032,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.559322033898305,
|
| 328 |
+
"grad_norm": 0.2560286819934845,
|
| 329 |
+
"learning_rate": 9.697183552209288e-05,
|
| 330 |
+
"loss": 0.004,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.5932203389830508,
|
| 335 |
+
"grad_norm": 0.17713002860546112,
|
| 336 |
+
"learning_rate": 9.677675760607734e-05,
|
| 337 |
+
"loss": 0.0024,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.6271186440677967,
|
| 342 |
+
"grad_norm": 0.14444805681705475,
|
| 343 |
+
"learning_rate": 9.657579953134383e-05,
|
| 344 |
+
"loss": 0.0035,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.6610169491525424,
|
| 349 |
+
"grad_norm": 0.26531147956848145,
|
| 350 |
+
"learning_rate": 9.636898655969837e-05,
|
| 351 |
+
"loss": 0.0023,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.694915254237288,
|
| 356 |
+
"grad_norm": 0.1393326222896576,
|
| 357 |
+
"learning_rate": 9.615634468894752e-05,
|
| 358 |
+
"loss": 0.001,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.7288135593220337,
|
| 363 |
+
"grad_norm": 0.22181938588619232,
|
| 364 |
+
"learning_rate": 9.593790064963046e-05,
|
| 365 |
+
"loss": 0.0018,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.7627118644067796,
|
| 370 |
+
"grad_norm": 0.1423530876636505,
|
| 371 |
+
"learning_rate": 9.571368190165863e-05,
|
| 372 |
+
"loss": 0.0029,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.7966101694915255,
|
| 377 |
+
"grad_norm": 0.1219254806637764,
|
| 378 |
+
"learning_rate": 9.548371663086392e-05,
|
| 379 |
+
"loss": 0.0022,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.8305084745762712,
|
| 384 |
+
"grad_norm": 0.1542597860097885,
|
| 385 |
+
"learning_rate": 9.524803374545548e-05,
|
| 386 |
+
"loss": 0.0035,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.8644067796610169,
|
| 391 |
+
"grad_norm": 0.16393965482711792,
|
| 392 |
+
"learning_rate": 9.500666287238573e-05,
|
| 393 |
+
"loss": 0.0021,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.8983050847457628,
|
| 398 |
+
"grad_norm": 0.35059654712677,
|
| 399 |
+
"learning_rate": 9.475963435362614e-05,
|
| 400 |
+
"loss": 0.0027,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.9322033898305084,
|
| 405 |
+
"grad_norm": 0.20335566997528076,
|
| 406 |
+
"learning_rate": 9.450697924235294e-05,
|
| 407 |
+
"loss": 0.0023,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.9661016949152543,
|
| 412 |
+
"grad_norm": 0.15571418404579163,
|
| 413 |
+
"learning_rate": 9.424872929904358e-05,
|
| 414 |
+
"loss": 0.0012,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 2.0,
|
| 419 |
+
"grad_norm": 0.29340890049934387,
|
| 420 |
+
"learning_rate": 9.398491698748422e-05,
|
| 421 |
+
"loss": 0.0022,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 2.0338983050847457,
|
| 426 |
+
"grad_norm": 0.2681172788143158,
|
| 427 |
+
"learning_rate": 9.371557547068878e-05,
|
| 428 |
+
"loss": 0.0013,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 2.0677966101694913,
|
| 433 |
+
"grad_norm": 0.16841526329517365,
|
| 434 |
+
"learning_rate": 9.344073860673015e-05,
|
| 435 |
+
"loss": 0.0012,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 2.1016949152542375,
|
| 440 |
+
"grad_norm": 0.07565703988075256,
|
| 441 |
+
"learning_rate": 9.316044094448392e-05,
|
| 442 |
+
"loss": 0.0013,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 2.135593220338983,
|
| 447 |
+
"grad_norm": 0.1882319450378418,
|
| 448 |
+
"learning_rate": 9.287471771928555e-05,
|
| 449 |
+
"loss": 0.0022,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.169491525423729,
|
| 454 |
+
"grad_norm": 0.07069575041532516,
|
| 455 |
+
"learning_rate": 9.25836048485008e-05,
|
| 456 |
+
"loss": 0.0007,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.2033898305084745,
|
| 461 |
+
"grad_norm": 0.16358712315559387,
|
| 462 |
+
"learning_rate": 9.228713892701077e-05,
|
| 463 |
+
"loss": 0.003,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.23728813559322,
|
| 468 |
+
"grad_norm": 0.13663357496261597,
|
| 469 |
+
"learning_rate": 9.198535722261181e-05,
|
| 470 |
+
"loss": 0.0019,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.2711864406779663,
|
| 475 |
+
"grad_norm": 0.13056139647960663,
|
| 476 |
+
"learning_rate": 9.167829767133047e-05,
|
| 477 |
+
"loss": 0.0013,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.305084745762712,
|
| 482 |
+
"grad_norm": 0.12309998273849487,
|
| 483 |
+
"learning_rate": 9.136599887265483e-05,
|
| 484 |
+
"loss": 0.0016,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.3389830508474576,
|
| 489 |
+
"grad_norm": 0.13602758944034576,
|
| 490 |
+
"learning_rate": 9.104850008468221e-05,
|
| 491 |
+
"loss": 0.0007,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.3728813559322033,
|
| 496 |
+
"grad_norm": 0.08036898076534271,
|
| 497 |
+
"learning_rate": 9.072584121918425e-05,
|
| 498 |
+
"loss": 0.0023,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.406779661016949,
|
| 503 |
+
"grad_norm": 0.08389495313167572,
|
| 504 |
+
"learning_rate": 9.039806283658962e-05,
|
| 505 |
+
"loss": 0.0014,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.440677966101695,
|
| 510 |
+
"grad_norm": 0.12567836046218872,
|
| 511 |
+
"learning_rate": 9.006520614088535e-05,
|
| 512 |
+
"loss": 0.001,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.4745762711864407,
|
| 517 |
+
"grad_norm": 0.15499231219291687,
|
| 518 |
+
"learning_rate": 8.972731297443722e-05,
|
| 519 |
+
"loss": 0.0018,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.5084745762711864,
|
| 524 |
+
"grad_norm": 0.09496314823627472,
|
| 525 |
+
"learning_rate": 8.938442581272983e-05,
|
| 526 |
+
"loss": 0.0031,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.542372881355932,
|
| 531 |
+
"grad_norm": 0.1453184187412262,
|
| 532 |
+
"learning_rate": 8.903658775902723e-05,
|
| 533 |
+
"loss": 0.0014,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.576271186440678,
|
| 538 |
+
"grad_norm": 0.11160401254892349,
|
| 539 |
+
"learning_rate": 8.868384253895445e-05,
|
| 540 |
+
"loss": 0.001,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.610169491525424,
|
| 545 |
+
"grad_norm": 0.10584841668605804,
|
| 546 |
+
"learning_rate": 8.832623449500095e-05,
|
| 547 |
+
"loss": 0.0016,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.6440677966101696,
|
| 552 |
+
"grad_norm": 0.094113789498806,
|
| 553 |
+
"learning_rate": 8.796380858094643e-05,
|
| 554 |
+
"loss": 0.0013,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.6779661016949152,
|
| 559 |
+
"grad_norm": 0.21183228492736816,
|
| 560 |
+
"learning_rate": 8.759661035620992e-05,
|
| 561 |
+
"loss": 0.0015,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.711864406779661,
|
| 566 |
+
"grad_norm": 0.14965838193893433,
|
| 567 |
+
"learning_rate": 8.722468598012245e-05,
|
| 568 |
+
"loss": 0.0018,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.7457627118644066,
|
| 573 |
+
"grad_norm": 0.08837207406759262,
|
| 574 |
+
"learning_rate": 8.684808220612478e-05,
|
| 575 |
+
"loss": 0.0008,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.7796610169491527,
|
| 580 |
+
"grad_norm": 0.11261831223964691,
|
| 581 |
+
"learning_rate": 8.646684637588991e-05,
|
| 582 |
+
"loss": 0.0022,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.8135593220338984,
|
| 587 |
+
"grad_norm": 0.221408873796463,
|
| 588 |
+
"learning_rate": 8.60810264133721e-05,
|
| 589 |
+
"loss": 0.0014,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.847457627118644,
|
| 594 |
+
"grad_norm": 0.37261417508125305,
|
| 595 |
+
"learning_rate": 8.56906708187824e-05,
|
| 596 |
+
"loss": 0.0023,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.8813559322033897,
|
| 601 |
+
"grad_norm": 0.12965640425682068,
|
| 602 |
+
"learning_rate": 8.529582866249186e-05,
|
| 603 |
+
"loss": 0.002,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.915254237288136,
|
| 608 |
+
"grad_norm": 0.12147118151187897,
|
| 609 |
+
"learning_rate": 8.489654957886306e-05,
|
| 610 |
+
"loss": 0.0015,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.9491525423728815,
|
| 615 |
+
"grad_norm": 0.13188952207565308,
|
| 616 |
+
"learning_rate": 8.449288376001075e-05,
|
| 617 |
+
"loss": 0.0013,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.983050847457627,
|
| 622 |
+
"grad_norm": 0.12438294291496277,
|
| 623 |
+
"learning_rate": 8.40848819494923e-05,
|
| 624 |
+
"loss": 0.0013,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 3.016949152542373,
|
| 629 |
+
"grad_norm": 0.309495747089386,
|
| 630 |
+
"learning_rate": 8.367259543592892e-05,
|
| 631 |
+
"loss": 0.0012,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 3.0508474576271185,
|
| 636 |
+
"grad_norm": 0.24989376962184906,
|
| 637 |
+
"learning_rate": 8.325607604655839e-05,
|
| 638 |
+
"loss": 0.001,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 3.084745762711864,
|
| 643 |
+
"grad_norm": 0.1818067878484726,
|
| 644 |
+
"learning_rate": 8.283537614071988e-05,
|
| 645 |
+
"loss": 0.001,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 3.1186440677966103,
|
| 650 |
+
"grad_norm": 0.13010039925575256,
|
| 651 |
+
"learning_rate": 8.241054860327216e-05,
|
| 652 |
+
"loss": 0.0013,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 3.152542372881356,
|
| 657 |
+
"grad_norm": 0.15505361557006836,
|
| 658 |
+
"learning_rate": 8.198164683794556e-05,
|
| 659 |
+
"loss": 0.0029,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 3.1864406779661016,
|
| 664 |
+
"grad_norm": 0.07065195590257645,
|
| 665 |
+
"learning_rate": 8.154872476062868e-05,
|
| 666 |
+
"loss": 0.0014,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 3.2203389830508473,
|
| 671 |
+
"grad_norm": 0.28758659958839417,
|
| 672 |
+
"learning_rate": 8.111183679259095e-05,
|
| 673 |
+
"loss": 0.002,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.2542372881355934,
|
| 678 |
+
"grad_norm": 0.18301035463809967,
|
| 679 |
+
"learning_rate": 8.067103785364139e-05,
|
| 680 |
+
"loss": 0.0014,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.288135593220339,
|
| 685 |
+
"grad_norm": 0.09719489514827728,
|
| 686 |
+
"learning_rate": 8.022638335522483e-05,
|
| 687 |
+
"loss": 0.0025,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.3220338983050848,
|
| 692 |
+
"grad_norm": 0.2065349668264389,
|
| 693 |
+
"learning_rate": 7.977792919345633e-05,
|
| 694 |
+
"loss": 0.0014,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.3559322033898304,
|
| 699 |
+
"grad_norm": 0.10643105208873749,
|
| 700 |
+
"learning_rate": 7.932573174209463e-05,
|
| 701 |
+
"loss": 0.0017,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.389830508474576,
|
| 706 |
+
"grad_norm": 0.11610566079616547,
|
| 707 |
+
"learning_rate": 7.886984784545566e-05,
|
| 708 |
+
"loss": 0.0008,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.423728813559322,
|
| 713 |
+
"grad_norm": 0.11425777524709702,
|
| 714 |
+
"learning_rate": 7.84103348112667e-05,
|
| 715 |
+
"loss": 0.0017,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.457627118644068,
|
| 720 |
+
"grad_norm": 0.08603143692016602,
|
| 721 |
+
"learning_rate": 7.794725040346251e-05,
|
| 722 |
+
"loss": 0.0011,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.4915254237288136,
|
| 727 |
+
"grad_norm": 0.14157338440418243,
|
| 728 |
+
"learning_rate": 7.748065283492396e-05,
|
| 729 |
+
"loss": 0.0019,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.5254237288135593,
|
| 734 |
+
"grad_norm": 0.3015165627002716,
|
| 735 |
+
"learning_rate": 7.701060076016024e-05,
|
| 736 |
+
"loss": 0.0016,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.559322033898305,
|
| 741 |
+
"grad_norm": 0.09349218755960464,
|
| 742 |
+
"learning_rate": 7.653715326793569e-05,
|
| 743 |
+
"loss": 0.0009,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.593220338983051,
|
| 748 |
+
"grad_norm": 0.13891376554965973,
|
| 749 |
+
"learning_rate": 7.606036987384184e-05,
|
| 750 |
+
"loss": 0.0009,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.6271186440677967,
|
| 755 |
+
"grad_norm": 0.10969799011945724,
|
| 756 |
+
"learning_rate": 7.558031051281592e-05,
|
| 757 |
+
"loss": 0.0016,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.6610169491525424,
|
| 762 |
+
"grad_norm": 0.16311612725257874,
|
| 763 |
+
"learning_rate": 7.509703553160666e-05,
|
| 764 |
+
"loss": 0.0015,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.694915254237288,
|
| 769 |
+
"grad_norm": 0.14392238855361938,
|
| 770 |
+
"learning_rate": 7.461060568118822e-05,
|
| 771 |
+
"loss": 0.0009,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.7288135593220337,
|
| 776 |
+
"grad_norm": 0.1457260549068451,
|
| 777 |
+
"learning_rate": 7.412108210912345e-05,
|
| 778 |
+
"loss": 0.0013,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.7627118644067794,
|
| 783 |
+
"grad_norm": 0.16048628091812134,
|
| 784 |
+
"learning_rate": 7.362852635187721e-05,
|
| 785 |
+
"loss": 0.0017,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.7966101694915255,
|
| 790 |
+
"grad_norm": 0.18469738960266113,
|
| 791 |
+
"learning_rate": 7.31330003270808e-05,
|
| 792 |
+
"loss": 0.0014,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.830508474576271,
|
| 797 |
+
"grad_norm": 0.180945485830307,
|
| 798 |
+
"learning_rate": 7.263456632574845e-05,
|
| 799 |
+
"loss": 0.0015,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.864406779661017,
|
| 804 |
+
"grad_norm": 0.14967080950737,
|
| 805 |
+
"learning_rate": 7.213328700444696e-05,
|
| 806 |
+
"loss": 0.0013,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.898305084745763,
|
| 811 |
+
"grad_norm": 0.25463271141052246,
|
| 812 |
+
"learning_rate": 7.162922537741937e-05,
|
| 813 |
+
"loss": 0.0023,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.9322033898305087,
|
| 818 |
+
"grad_norm": 0.24937133491039276,
|
| 819 |
+
"learning_rate": 7.112244480866356e-05,
|
| 820 |
+
"loss": 0.0018,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.9661016949152543,
|
| 825 |
+
"grad_norm": 0.1597055047750473,
|
| 826 |
+
"learning_rate": 7.061300900396703e-05,
|
| 827 |
+
"loss": 0.0032,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 4.0,
|
| 832 |
+
"grad_norm": 0.14016273617744446,
|
| 833 |
+
"learning_rate": 7.010098200289859e-05,
|
| 834 |
+
"loss": 0.0014,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 4.033898305084746,
|
| 839 |
+
"grad_norm": 0.11648664623498917,
|
| 840 |
+
"learning_rate": 6.958642817075823e-05,
|
| 841 |
+
"loss": 0.001,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 4.067796610169491,
|
| 846 |
+
"grad_norm": 0.23266953229904175,
|
| 847 |
+
"learning_rate": 6.906941219048584e-05,
|
| 848 |
+
"loss": 0.001,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 4.101694915254237,
|
| 853 |
+
"grad_norm": 0.11199619621038437,
|
| 854 |
+
"learning_rate": 6.854999905453021e-05,
|
| 855 |
+
"loss": 0.0008,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 4.135593220338983,
|
| 860 |
+
"grad_norm": 0.09246116131544113,
|
| 861 |
+
"learning_rate": 6.802825405667905e-05,
|
| 862 |
+
"loss": 0.001,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 4.169491525423728,
|
| 867 |
+
"grad_norm": 0.08600598573684692,
|
| 868 |
+
"learning_rate": 6.750424278385103e-05,
|
| 869 |
+
"loss": 0.0008,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 4.203389830508475,
|
| 874 |
+
"grad_norm": 0.09983401745557785,
|
| 875 |
+
"learning_rate": 6.697803110785115e-05,
|
| 876 |
+
"loss": 0.0009,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 4.237288135593221,
|
| 881 |
+
"grad_norm": 0.09594856947660446,
|
| 882 |
+
"learning_rate": 6.644968517709009e-05,
|
| 883 |
+
"loss": 0.0015,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.271186440677966,
|
| 888 |
+
"grad_norm": 0.08599961549043655,
|
| 889 |
+
"learning_rate": 6.591927140826902e-05,
|
| 890 |
+
"loss": 0.0007,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.305084745762712,
|
| 895 |
+
"grad_norm": 0.16801735758781433,
|
| 896 |
+
"learning_rate": 6.538685647803049e-05,
|
| 897 |
+
"loss": 0.001,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 4.338983050847458,
|
| 902 |
+
"grad_norm": 0.04569881409406662,
|
| 903 |
+
"learning_rate": 6.485250731457678e-05,
|
| 904 |
+
"loss": 0.0007,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.372881355932203,
|
| 909 |
+
"grad_norm": 0.04355693981051445,
|
| 910 |
+
"learning_rate": 6.431629108925646e-05,
|
| 911 |
+
"loss": 0.0006,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.406779661016949,
|
| 916 |
+
"grad_norm": 0.1598712056875229,
|
| 917 |
+
"learning_rate": 6.377827520812061e-05,
|
| 918 |
+
"loss": 0.0013,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.440677966101695,
|
| 923 |
+
"grad_norm": 0.1439727395772934,
|
| 924 |
+
"learning_rate": 6.323852730344935e-05,
|
| 925 |
+
"loss": 0.0006,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.47457627118644,
|
| 930 |
+
"grad_norm": 0.07059933245182037,
|
| 931 |
+
"learning_rate": 6.269711522525006e-05,
|
| 932 |
+
"loss": 0.0004,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.508474576271187,
|
| 937 |
+
"grad_norm": 0.2104925960302353,
|
| 938 |
+
"learning_rate": 6.215410703272805e-05,
|
| 939 |
+
"loss": 0.0019,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.5423728813559325,
|
| 944 |
+
"grad_norm": 0.3595544993877411,
|
| 945 |
+
"learning_rate": 6.160957098573119e-05,
|
| 946 |
+
"loss": 0.0026,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.576271186440678,
|
| 951 |
+
"grad_norm": 0.09160851687192917,
|
| 952 |
+
"learning_rate": 6.10635755361691e-05,
|
| 953 |
+
"loss": 0.0012,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.610169491525424,
|
| 958 |
+
"grad_norm": 0.15141859650611877,
|
| 959 |
+
"learning_rate": 6.05161893194083e-05,
|
| 960 |
+
"loss": 0.0012,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.6440677966101696,
|
| 965 |
+
"grad_norm": 0.24170026183128357,
|
| 966 |
+
"learning_rate": 5.9967481145644265e-05,
|
| 967 |
+
"loss": 0.0009,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.677966101694915,
|
| 972 |
+
"grad_norm": 0.12173440307378769,
|
| 973 |
+
"learning_rate": 5.941751999125149e-05,
|
| 974 |
+
"loss": 0.0009,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.711864406779661,
|
| 979 |
+
"grad_norm": 0.2746393084526062,
|
| 980 |
+
"learning_rate": 5.886637499011278e-05,
|
| 981 |
+
"loss": 0.0008,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.745762711864407,
|
| 986 |
+
"grad_norm": 0.06491386145353317,
|
| 987 |
+
"learning_rate": 5.831411542492854e-05,
|
| 988 |
+
"loss": 0.0006,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.779661016949152,
|
| 993 |
+
"grad_norm": 0.10497017949819565,
|
| 994 |
+
"learning_rate": 5.7760810718507487e-05,
|
| 995 |
+
"loss": 0.0013,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.813559322033898,
|
| 1000 |
+
"grad_norm": 0.08166036754846573,
|
| 1001 |
+
"learning_rate": 5.720653042503978e-05,
|
| 1002 |
+
"loss": 0.0007,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.847457627118644,
|
| 1007 |
+
"grad_norm": 0.08851595968008041,
|
| 1008 |
+
"learning_rate": 5.665134422135354e-05,
|
| 1009 |
+
"loss": 0.0011,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.88135593220339,
|
| 1014 |
+
"grad_norm": 0.07782018929719925,
|
| 1015 |
+
"learning_rate": 5.6095321898156016e-05,
|
| 1016 |
+
"loss": 0.0014,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.915254237288136,
|
| 1021 |
+
"grad_norm": 0.06927744299173355,
|
| 1022 |
+
"learning_rate": 5.553853335126039e-05,
|
| 1023 |
+
"loss": 0.0004,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.9491525423728815,
|
| 1028 |
+
"grad_norm": 0.09779927879571915,
|
| 1029 |
+
"learning_rate": 5.498104857279941e-05,
|
| 1030 |
+
"loss": 0.0006,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.983050847457627,
|
| 1035 |
+
"grad_norm": 0.18358203768730164,
|
| 1036 |
+
"learning_rate": 5.4422937642426906e-05,
|
| 1037 |
+
"loss": 0.0009,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 5.016949152542373,
|
| 1042 |
+
"grad_norm": 0.1410626918077469,
|
| 1043 |
+
"learning_rate": 5.3864270718508305e-05,
|
| 1044 |
+
"loss": 0.0009,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 5.0508474576271185,
|
| 1049 |
+
"grad_norm": 0.17358587682247162,
|
| 1050 |
+
"learning_rate": 5.330511802930117e-05,
|
| 1051 |
+
"loss": 0.0011,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 5.084745762711864,
|
| 1056 |
+
"grad_norm": 0.1470521092414856,
|
| 1057 |
+
"learning_rate": 5.274554986412716e-05,
|
| 1058 |
+
"loss": 0.0011,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 5.11864406779661,
|
| 1063 |
+
"grad_norm": 0.1285671442747116,
|
| 1064 |
+
"learning_rate": 5.218563656453609e-05,
|
| 1065 |
+
"loss": 0.0012,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 5.1525423728813555,
|
| 1070 |
+
"grad_norm": 0.10445746779441833,
|
| 1071 |
+
"learning_rate": 5.162544851546349e-05,
|
| 1072 |
+
"loss": 0.0017,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 5.186440677966102,
|
| 1077 |
+
"grad_norm": 0.14065216481685638,
|
| 1078 |
+
"learning_rate": 5.106505613638284e-05,
|
| 1079 |
+
"loss": 0.0022,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 5.220338983050848,
|
| 1084 |
+
"grad_norm": 0.09856051951646805,
|
| 1085 |
+
"learning_rate": 5.0504529872453256e-05,
|
| 1086 |
+
"loss": 0.002,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 5.254237288135593,
|
| 1091 |
+
"grad_norm": 0.1747041493654251,
|
| 1092 |
+
"learning_rate": 4.9943940185664124e-05,
|
| 1093 |
+
"loss": 0.001,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 5.288135593220339,
|
| 1098 |
+
"grad_norm": 0.08963986486196518,
|
| 1099 |
+
"learning_rate": 4.9383357545977497e-05,
|
| 1100 |
+
"loss": 0.0007,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 5.322033898305085,
|
| 1105 |
+
"grad_norm": 0.09163827449083328,
|
| 1106 |
+
"learning_rate": 4.882285242246957e-05,
|
| 1107 |
+
"loss": 0.001,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 5.3559322033898304,
|
| 1112 |
+
"grad_norm": 0.08859282732009888,
|
| 1113 |
+
"learning_rate": 4.8262495274472225e-05,
|
| 1114 |
+
"loss": 0.0006,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 5.389830508474576,
|
| 1119 |
+
"grad_norm": 0.06251803785562515,
|
| 1120 |
+
"learning_rate": 4.770235654271582e-05,
|
| 1121 |
+
"loss": 0.0004,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 5.423728813559322,
|
| 1126 |
+
"grad_norm": 0.06271404027938843,
|
| 1127 |
+
"learning_rate": 4.7142506640474274e-05,
|
| 1128 |
+
"loss": 0.0005,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 5.4576271186440675,
|
| 1133 |
+
"grad_norm": 0.16893021762371063,
|
| 1134 |
+
"learning_rate": 4.658301594471367e-05,
|
| 1135 |
+
"loss": 0.0009,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 5.491525423728813,
|
| 1140 |
+
"grad_norm": 0.04200820252299309,
|
| 1141 |
+
"learning_rate": 4.602395478724539e-05,
|
| 1142 |
+
"loss": 0.0006,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 5.52542372881356,
|
| 1147 |
+
"grad_norm": 0.030414093285799026,
|
| 1148 |
+
"learning_rate": 4.546539344588486e-05,
|
| 1149 |
+
"loss": 0.0005,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.559322033898305,
|
| 1154 |
+
"grad_norm": 0.10116266459226608,
|
| 1155 |
+
"learning_rate": 4.490740213561727e-05,
|
| 1156 |
+
"loss": 0.0011,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.593220338983051,
|
| 1161 |
+
"grad_norm": 0.12049903720617294,
|
| 1162 |
+
"learning_rate": 4.435005099977093e-05,
|
| 1163 |
+
"loss": 0.0006,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.627118644067797,
|
| 1168 |
+
"grad_norm": 0.0699400007724762,
|
| 1169 |
+
"learning_rate": 4.379341010119992e-05,
|
| 1170 |
+
"loss": 0.0005,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.661016949152542,
|
| 1175 |
+
"grad_norm": 0.06926112622022629,
|
| 1176 |
+
"learning_rate": 4.323754941347654e-05,
|
| 1177 |
+
"loss": 0.0006,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 5.694915254237288,
|
| 1182 |
+
"grad_norm": 0.06657879799604416,
|
| 1183 |
+
"learning_rate": 4.268253881209532e-05,
|
| 1184 |
+
"loss": 0.0005,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.728813559322034,
|
| 1189 |
+
"grad_norm": 0.04170970991253853,
|
| 1190 |
+
"learning_rate": 4.2128448065689065e-05,
|
| 1191 |
+
"loss": 0.0005,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.762711864406779,
|
| 1196 |
+
"grad_norm": 0.09546169638633728,
|
| 1197 |
+
"learning_rate": 4.157534682725856e-05,
|
| 1198 |
+
"loss": 0.0004,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 5.796610169491525,
|
| 1203 |
+
"grad_norm": 0.14321397244930267,
|
| 1204 |
+
"learning_rate": 4.102330462541662e-05,
|
| 1205 |
+
"loss": 0.0005,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.830508474576272,
|
| 1210 |
+
"grad_norm": 0.06389990448951721,
|
| 1211 |
+
"learning_rate": 4.047239085564794e-05,
|
| 1212 |
+
"loss": 0.0005,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.864406779661017,
|
| 1217 |
+
"grad_norm": 0.03475102782249451,
|
| 1218 |
+
"learning_rate": 3.9922674771585546e-05,
|
| 1219 |
+
"loss": 0.0003,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 5.898305084745763,
|
| 1224 |
+
"grad_norm": 0.08064165711402893,
|
| 1225 |
+
"learning_rate": 3.937422547630519e-05,
|
| 1226 |
+
"loss": 0.0015,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 5.932203389830509,
|
| 1231 |
+
"grad_norm": 0.05471855029463768,
|
| 1232 |
+
"learning_rate": 3.882711191363859e-05,
|
| 1233 |
+
"loss": 0.0008,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.966101694915254,
|
| 1238 |
+
"grad_norm": 0.25674968957901,
|
| 1239 |
+
"learning_rate": 3.828140285950676e-05,
|
| 1240 |
+
"loss": 0.0015,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 6.0,
|
| 1245 |
+
"grad_norm": 0.03591851517558098,
|
| 1246 |
+
"learning_rate": 3.773716691327432e-05,
|
| 1247 |
+
"loss": 0.0007,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 6.033898305084746,
|
| 1252 |
+
"grad_norm": 0.228873610496521,
|
| 1253 |
+
"learning_rate": 3.7194472489126176e-05,
|
| 1254 |
+
"loss": 0.001,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 6.067796610169491,
|
| 1259 |
+
"grad_norm": 0.10522802919149399,
|
| 1260 |
+
"learning_rate": 3.665338780746736e-05,
|
| 1261 |
+
"loss": 0.0007,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 6.101694915254237,
|
| 1266 |
+
"grad_norm": 0.1092459186911583,
|
| 1267 |
+
"learning_rate": 3.611398088634721e-05,
|
| 1268 |
+
"loss": 0.0004,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 6.135593220338983,
|
| 1273 |
+
"grad_norm": 0.15313047170639038,
|
| 1274 |
+
"learning_rate": 3.5576319532909145e-05,
|
| 1275 |
+
"loss": 0.0004,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 6.169491525423728,
|
| 1280 |
+
"grad_norm": 0.04173111170530319,
|
| 1281 |
+
"learning_rate": 3.5040471334866695e-05,
|
| 1282 |
+
"loss": 0.0004,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 6.203389830508475,
|
| 1287 |
+
"grad_norm": 0.02953781560063362,
|
| 1288 |
+
"learning_rate": 3.450650365200741e-05,
|
| 1289 |
+
"loss": 0.0003,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 6.237288135593221,
|
| 1294 |
+
"grad_norm": 0.04410599172115326,
|
| 1295 |
+
"learning_rate": 3.397448360772516e-05,
|
| 1296 |
+
"loss": 0.0006,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 6.271186440677966,
|
| 1301 |
+
"grad_norm": 0.09171520918607712,
|
| 1302 |
+
"learning_rate": 3.344447808058232e-05,
|
| 1303 |
+
"loss": 0.0005,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 6.305084745762712,
|
| 1308 |
+
"grad_norm": 0.051552463322877884,
|
| 1309 |
+
"learning_rate": 3.291655369590269e-05,
|
| 1310 |
+
"loss": 0.0006,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 6.338983050847458,
|
| 1315 |
+
"grad_norm": 0.03626188263297081,
|
| 1316 |
+
"learning_rate": 3.239077681739618e-05,
|
| 1317 |
+
"loss": 0.0005,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 6.372881355932203,
|
| 1322 |
+
"grad_norm": 0.09546960145235062,
|
| 1323 |
+
"learning_rate": 3.186721353881648e-05,
|
| 1324 |
+
"loss": 0.0004,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 6.406779661016949,
|
| 1329 |
+
"grad_norm": 0.0609428696334362,
|
| 1330 |
+
"learning_rate": 3.134592967565263e-05,
|
| 1331 |
+
"loss": 0.0003,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 6.440677966101695,
|
| 1336 |
+
"grad_norm": 0.05546706169843674,
|
| 1337 |
+
"learning_rate": 3.082699075685553e-05,
|
| 1338 |
+
"loss": 0.0003,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 6.47457627118644,
|
| 1343 |
+
"grad_norm": 0.03381851688027382,
|
| 1344 |
+
"learning_rate": 3.0310462016600504e-05,
|
| 1345 |
+
"loss": 0.0006,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 6.508474576271187,
|
| 1350 |
+
"grad_norm": 0.0640668272972107,
|
| 1351 |
+
"learning_rate": 2.9796408386086962e-05,
|
| 1352 |
+
"loss": 0.0003,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 6.5423728813559325,
|
| 1357 |
+
"grad_norm": 0.07631280273199081,
|
| 1358 |
+
"learning_rate": 2.9284894485376057e-05,
|
| 1359 |
+
"loss": 0.0004,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 6.576271186440678,
|
| 1364 |
+
"grad_norm": 0.1757967323064804,
|
| 1365 |
+
"learning_rate": 2.8775984615267504e-05,
|
| 1366 |
+
"loss": 0.0007,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 6.610169491525424,
|
| 1371 |
+
"grad_norm": 0.05827037990093231,
|
| 1372 |
+
"learning_rate": 2.8269742749216566e-05,
|
| 1373 |
+
"loss": 0.0005,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 6.6440677966101696,
|
| 1378 |
+
"grad_norm": 0.07589293271303177,
|
| 1379 |
+
"learning_rate": 2.7766232525292103e-05,
|
| 1380 |
+
"loss": 0.0003,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 6.677966101694915,
|
| 1385 |
+
"grad_norm": 0.0697827935218811,
|
| 1386 |
+
"learning_rate": 2.7265517238176847e-05,
|
| 1387 |
+
"loss": 0.0008,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 6.711864406779661,
|
| 1392 |
+
"grad_norm": 0.11972369253635406,
|
| 1393 |
+
"learning_rate": 2.676765983121089e-05,
|
| 1394 |
+
"loss": 0.0005,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 6.745762711864407,
|
| 1399 |
+
"grad_norm": 0.0417722649872303,
|
| 1400 |
+
"learning_rate": 2.627272288847915e-05,
|
| 1401 |
+
"loss": 0.0003,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 6.779661016949152,
|
| 1406 |
+
"grad_norm": 0.04584512487053871,
|
| 1407 |
+
"learning_rate": 2.578076862694426e-05,
|
| 1408 |
+
"loss": 0.0007,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 6.813559322033898,
|
| 1413 |
+
"grad_norm": 0.07800233364105225,
|
| 1414 |
+
"learning_rate": 2.5291858888625368e-05,
|
| 1415 |
+
"loss": 0.0004,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 6.847457627118644,
|
| 1420 |
+
"grad_norm": 0.07211495190858841,
|
| 1421 |
+
"learning_rate": 2.4806055132824185e-05,
|
| 1422 |
+
"loss": 0.0003,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 6.88135593220339,
|
| 1427 |
+
"grad_norm": 0.06364759802818298,
|
| 1428 |
+
"learning_rate": 2.4323418428399203e-05,
|
| 1429 |
+
"loss": 0.0003,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 6.915254237288136,
|
| 1434 |
+
"grad_norm": 0.07082319259643555,
|
| 1435 |
+
"learning_rate": 2.384400944608886e-05,
|
| 1436 |
+
"loss": 0.0007,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 6.9491525423728815,
|
| 1441 |
+
"grad_norm": 0.06931556761264801,
|
| 1442 |
+
"learning_rate": 2.336788845088478e-05,
|
| 1443 |
+
"loss": 0.0004,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 6.983050847457627,
|
| 1448 |
+
"grad_norm": 0.15574094653129578,
|
| 1449 |
+
"learning_rate": 2.289511529445616e-05,
|
| 1450 |
+
"loss": 0.0006,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 7.016949152542373,
|
| 1455 |
+
"grad_norm": 0.04960203543305397,
|
| 1456 |
+
"learning_rate": 2.242574940762589e-05,
|
| 1457 |
+
"loss": 0.0003,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 7.0508474576271185,
|
| 1462 |
+
"grad_norm": 0.049310192465782166,
|
| 1463 |
+
"learning_rate": 2.195984979289974e-05,
|
| 1464 |
+
"loss": 0.0005,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 7.084745762711864,
|
| 1469 |
+
"grad_norm": 0.06411266326904297,
|
| 1470 |
+
"learning_rate": 2.149747501704939e-05,
|
| 1471 |
+
"loss": 0.0004,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 7.11864406779661,
|
| 1476 |
+
"grad_norm": 0.06982172280550003,
|
| 1477 |
+
"learning_rate": 2.1038683203750092e-05,
|
| 1478 |
+
"loss": 0.0005,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 7.1525423728813555,
|
| 1483 |
+
"grad_norm": 0.11477413028478622,
|
| 1484 |
+
"learning_rate": 2.058353202627417e-05,
|
| 1485 |
+
"loss": 0.0004,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 7.186440677966102,
|
| 1490 |
+
"grad_norm": 0.05071744695305824,
|
| 1491 |
+
"learning_rate": 2.0132078700241158e-05,
|
| 1492 |
+
"loss": 0.0003,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 7.220338983050848,
|
| 1497 |
+
"grad_norm": 0.06909611821174622,
|
| 1498 |
+
"learning_rate": 1.968437997642526e-05,
|
| 1499 |
+
"loss": 0.0004,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 7.254237288135593,
|
| 1504 |
+
"grad_norm": 0.08218582719564438,
|
| 1505 |
+
"learning_rate": 1.924049213362153e-05,
|
| 1506 |
+
"loss": 0.0005,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 7.288135593220339,
|
| 1511 |
+
"grad_norm": 0.10386521369218826,
|
| 1512 |
+
"learning_rate": 1.8800470971571184e-05,
|
| 1513 |
+
"loss": 0.0003,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 7.322033898305085,
|
| 1518 |
+
"grad_norm": 0.052584659308195114,
|
| 1519 |
+
"learning_rate": 1.836437180394715e-05,
|
| 1520 |
+
"loss": 0.0002,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 7.3559322033898304,
|
| 1525 |
+
"grad_norm": 0.0438753217458725,
|
| 1526 |
+
"learning_rate": 1.793224945140086e-05,
|
| 1527 |
+
"loss": 0.0003,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 7.389830508474576,
|
| 1532 |
+
"grad_norm": 0.04817179590463638,
|
| 1533 |
+
"learning_rate": 1.750415823467082e-05,
|
| 1534 |
+
"loss": 0.0002,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 7.423728813559322,
|
| 1539 |
+
"grad_norm": 0.09476613253355026,
|
| 1540 |
+
"learning_rate": 1.7080151967754142e-05,
|
| 1541 |
+
"loss": 0.0004,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 7.4576271186440675,
|
| 1546 |
+
"grad_norm": 0.04253845289349556,
|
| 1547 |
+
"learning_rate": 1.666028395114185e-05,
|
| 1548 |
+
"loss": 0.0004,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 7.491525423728813,
|
| 1553 |
+
"grad_norm": 0.055023740977048874,
|
| 1554 |
+
"learning_rate": 1.62446069651185e-05,
|
| 1555 |
+
"loss": 0.0009,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 7.52542372881356,
|
| 1560 |
+
"grad_norm": 0.11626702547073364,
|
| 1561 |
+
"learning_rate": 1.5833173263127426e-05,
|
| 1562 |
+
"loss": 0.0007,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 7.559322033898305,
|
| 1567 |
+
"grad_norm": 0.1153387799859047,
|
| 1568 |
+
"learning_rate": 1.5426034565202142e-05,
|
| 1569 |
+
"loss": 0.0007,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 7.593220338983051,
|
| 1574 |
+
"grad_norm": 0.07169368118047714,
|
| 1575 |
+
"learning_rate": 1.5023242051464675e-05,
|
| 1576 |
+
"loss": 0.0004,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 7.627118644067797,
|
| 1581 |
+
"grad_norm": 0.061480406671762466,
|
| 1582 |
+
"learning_rate": 1.4624846355692001e-05,
|
| 1583 |
+
"loss": 0.0004,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 7.661016949152542,
|
| 1588 |
+
"grad_norm": 0.0765593945980072,
|
| 1589 |
+
"learning_rate": 1.4230897558950951e-05,
|
| 1590 |
+
"loss": 0.001,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 7.694915254237288,
|
| 1595 |
+
"grad_norm": 0.06186467036604881,
|
| 1596 |
+
"learning_rate": 1.384144518330268e-05,
|
| 1597 |
+
"loss": 0.0003,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 7.728813559322034,
|
| 1602 |
+
"grad_norm": 0.02641032636165619,
|
| 1603 |
+
"learning_rate": 1.3456538185577466e-05,
|
| 1604 |
+
"loss": 0.0003,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 7.762711864406779,
|
| 1609 |
+
"grad_norm": 0.08123563975095749,
|
| 1610 |
+
"learning_rate": 1.3076224951220411e-05,
|
| 1611 |
+
"loss": 0.0009,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 7.796610169491525,
|
| 1616 |
+
"grad_norm": 0.09227631241083145,
|
| 1617 |
+
"learning_rate": 1.2700553288209126e-05,
|
| 1618 |
+
"loss": 0.0003,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 7.830508474576272,
|
| 1623 |
+
"grad_norm": 0.03893280774354935,
|
| 1624 |
+
"learning_rate": 1.2329570421043957e-05,
|
| 1625 |
+
"loss": 0.0003,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 7.864406779661017,
|
| 1630 |
+
"grad_norm": 0.04822898656129837,
|
| 1631 |
+
"learning_rate": 1.1963322984811453e-05,
|
| 1632 |
+
"loss": 0.0008,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 7.898305084745763,
|
| 1637 |
+
"grad_norm": 0.05023796483874321,
|
| 1638 |
+
"learning_rate": 1.1601857019322093e-05,
|
| 1639 |
+
"loss": 0.0003,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 7.932203389830509,
|
| 1644 |
+
"grad_norm": 0.09390699863433838,
|
| 1645 |
+
"learning_rate": 1.1245217963322763e-05,
|
| 1646 |
+
"loss": 0.0003,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 7.966101694915254,
|
| 1651 |
+
"grad_norm": 0.016306394711136818,
|
| 1652 |
+
"learning_rate": 1.0893450648784737e-05,
|
| 1653 |
+
"loss": 0.002,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 8.0,
|
| 1658 |
+
"grad_norm": 0.04842328652739525,
|
| 1659 |
+
"learning_rate": 1.0546599295268061e-05,
|
| 1660 |
+
"loss": 0.0003,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 8.033898305084746,
|
| 1665 |
+
"grad_norm": 0.034694500267505646,
|
| 1666 |
+
"learning_rate": 1.0204707504362826e-05,
|
| 1667 |
+
"loss": 0.0003,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 8.067796610169491,
|
| 1672 |
+
"grad_norm": 0.049791429191827774,
|
| 1673 |
+
"learning_rate": 9.867818254208123e-06,
|
| 1674 |
+
"loss": 0.0007,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 8.101694915254237,
|
| 1679 |
+
"grad_norm": 0.02616948075592518,
|
| 1680 |
+
"learning_rate": 9.535973894089401e-06,
|
| 1681 |
+
"loss": 0.0004,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 8.135593220338983,
|
| 1686 |
+
"grad_norm": 0.08042062073945999,
|
| 1687 |
+
"learning_rate": 9.209216139114935e-06,
|
| 1688 |
+
"loss": 0.0002,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 8.169491525423728,
|
| 1693 |
+
"grad_norm": 0.03833793103694916,
|
| 1694 |
+
"learning_rate": 8.887586064971858e-06,
|
| 1695 |
+
"loss": 0.0004,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 8.203389830508474,
|
| 1700 |
+
"grad_norm": 0.04440383240580559,
|
| 1701 |
+
"learning_rate": 8.571124102762768e-06,
|
| 1702 |
+
"loss": 0.0003,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 8.23728813559322,
|
| 1707 |
+
"grad_norm": 0.026959413662552834,
|
| 1708 |
+
"learning_rate": 8.259870033923151e-06,
|
| 1709 |
+
"loss": 0.0012,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 8.271186440677965,
|
| 1714 |
+
"grad_norm": 0.037755124270915985,
|
| 1715 |
+
"learning_rate": 7.95386298522065e-06,
|
| 1716 |
+
"loss": 0.0004,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 8.305084745762711,
|
| 1721 |
+
"grad_norm": 0.1117471233010292,
|
| 1722 |
+
"learning_rate": 7.653141423836584e-06,
|
| 1723 |
+
"loss": 0.0011,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 8.338983050847457,
|
| 1728 |
+
"grad_norm": 0.09815044701099396,
|
| 1729 |
+
"learning_rate": 7.357743152530272e-06,
|
| 1730 |
+
"loss": 0.0006,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 8.372881355932204,
|
| 1735 |
+
"grad_norm": 0.09394808858633041,
|
| 1736 |
+
"learning_rate": 7.067705304887074e-06,
|
| 1737 |
+
"loss": 0.0005,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 8.40677966101695,
|
| 1742 |
+
"grad_norm": 0.06960106641054153,
|
| 1743 |
+
"learning_rate": 6.783064340650364e-06,
|
| 1744 |
+
"loss": 0.0003,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 8.440677966101696,
|
| 1749 |
+
"grad_norm": 0.035113442689180374,
|
| 1750 |
+
"learning_rate": 6.503856041138323e-06,
|
| 1751 |
+
"loss": 0.0005,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 8.474576271186441,
|
| 1756 |
+
"grad_norm": 0.02795569784939289,
|
| 1757 |
+
"learning_rate": 6.230115504745953e-06,
|
| 1758 |
+
"loss": 0.0012,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 8.508474576271187,
|
| 1763 |
+
"grad_norm": 0.02803714945912361,
|
| 1764 |
+
"learning_rate": 5.961877142533056e-06,
|
| 1765 |
+
"loss": 0.0004,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 8.542372881355933,
|
| 1770 |
+
"grad_norm": 0.009161763824522495,
|
| 1771 |
+
"learning_rate": 5.699174673898394e-06,
|
| 1772 |
+
"loss": 0.0003,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 8.576271186440678,
|
| 1777 |
+
"grad_norm": 0.026831788942217827,
|
| 1778 |
+
"learning_rate": 5.4420411223410574e-06,
|
| 1779 |
+
"loss": 0.0004,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 8.610169491525424,
|
| 1784 |
+
"grad_norm": 0.011004606261849403,
|
| 1785 |
+
"learning_rate": 5.190508811309091e-06,
|
| 1786 |
+
"loss": 0.0002,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 8.64406779661017,
|
| 1791 |
+
"grad_norm": 0.027460021898150444,
|
| 1792 |
+
"learning_rate": 4.944609360136271e-06,
|
| 1793 |
+
"loss": 0.0002,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 8.677966101694915,
|
| 1798 |
+
"grad_norm": 0.04089871421456337,
|
| 1799 |
+
"learning_rate": 4.704373680067325e-06,
|
| 1800 |
+
"loss": 0.0002,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 8.711864406779661,
|
| 1805 |
+
"grad_norm": 0.03699534758925438,
|
| 1806 |
+
"learning_rate": 4.469831970372146e-06,
|
| 1807 |
+
"loss": 0.0002,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 8.745762711864407,
|
| 1812 |
+
"grad_norm": 0.021349631249904633,
|
| 1813 |
+
"learning_rate": 4.241013714549597e-06,
|
| 1814 |
+
"loss": 0.0002,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 8.779661016949152,
|
| 1819 |
+
"grad_norm": 0.02745555341243744,
|
| 1820 |
+
"learning_rate": 4.017947676621187e-06,
|
| 1821 |
+
"loss": 0.0004,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 8.813559322033898,
|
| 1826 |
+
"grad_norm": 0.013199915178120136,
|
| 1827 |
+
"learning_rate": 3.800661897515245e-06,
|
| 1828 |
+
"loss": 0.0002,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 8.847457627118644,
|
| 1833 |
+
"grad_norm": 0.026064835488796234,
|
| 1834 |
+
"learning_rate": 3.589183691541981e-06,
|
| 1835 |
+
"loss": 0.0002,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 8.88135593220339,
|
| 1840 |
+
"grad_norm": 0.01884126476943493,
|
| 1841 |
+
"learning_rate": 3.3835396429599152e-06,
|
| 1842 |
+
"loss": 0.0003,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 8.915254237288135,
|
| 1847 |
+
"grad_norm": 0.03649914637207985,
|
| 1848 |
+
"learning_rate": 3.18375560263402e-06,
|
| 1849 |
+
"loss": 0.0002,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 8.94915254237288,
|
| 1854 |
+
"grad_norm": 0.020875606685876846,
|
| 1855 |
+
"learning_rate": 2.9898566847861245e-06,
|
| 1856 |
+
"loss": 0.0003,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 8.983050847457626,
|
| 1861 |
+
"grad_norm": 0.05111026391386986,
|
| 1862 |
+
"learning_rate": 2.8018672638378483e-06,
|
| 1863 |
+
"loss": 0.0002,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 9.016949152542374,
|
| 1868 |
+
"grad_norm": 0.017365949228405952,
|
| 1869 |
+
"learning_rate": 2.619810971346587e-06,
|
| 1870 |
+
"loss": 0.0003,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 9.05084745762712,
|
| 1875 |
+
"grad_norm": 0.09889742732048035,
|
| 1876 |
+
"learning_rate": 2.443710693034884e-06,
|
| 1877 |
+
"loss": 0.0003,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 9.084745762711865,
|
| 1882 |
+
"grad_norm": 0.029447954148054123,
|
| 1883 |
+
"learning_rate": 2.2735885659134925e-06,
|
| 1884 |
+
"loss": 0.0003,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 9.11864406779661,
|
| 1889 |
+
"grad_norm": 0.0224969033151865,
|
| 1890 |
+
"learning_rate": 2.1094659754986477e-06,
|
| 1891 |
+
"loss": 0.0002,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 9.152542372881356,
|
| 1896 |
+
"grad_norm": 0.010688988491892815,
|
| 1897 |
+
"learning_rate": 1.9513635531237417e-06,
|
| 1898 |
+
"loss": 0.0003,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 9.186440677966102,
|
| 1903 |
+
"grad_norm": 0.059495627880096436,
|
| 1904 |
+
"learning_rate": 1.7993011733458075e-06,
|
| 1905 |
+
"loss": 0.0003,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 9.220338983050848,
|
| 1910 |
+
"grad_norm": 0.007177701219916344,
|
| 1911 |
+
"learning_rate": 1.6532979514471747e-06,
|
| 1912 |
+
"loss": 0.0002,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 9.254237288135593,
|
| 1917 |
+
"grad_norm": 0.052893899381160736,
|
| 1918 |
+
"learning_rate": 1.5133722410325457e-06,
|
| 1919 |
+
"loss": 0.0008,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 9.288135593220339,
|
| 1924 |
+
"grad_norm": 0.012546384707093239,
|
| 1925 |
+
"learning_rate": 1.3795416317218036e-06,
|
| 1926 |
+
"loss": 0.0007,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 9.322033898305085,
|
| 1931 |
+
"grad_norm": 0.008967459201812744,
|
| 1932 |
+
"learning_rate": 1.2518229469388965e-06,
|
| 1933 |
+
"loss": 0.0002,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 9.35593220338983,
|
| 1938 |
+
"grad_norm": 0.11077430099248886,
|
| 1939 |
+
"learning_rate": 1.1302322417970135e-06,
|
| 1940 |
+
"loss": 0.0002,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 9.389830508474576,
|
| 1945 |
+
"grad_norm": 0.014416457153856754,
|
| 1946 |
+
"learning_rate": 1.0147848010803318e-06,
|
| 1947 |
+
"loss": 0.0001,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 9.423728813559322,
|
| 1952 |
+
"grad_norm": 0.016251103952527046,
|
| 1953 |
+
"learning_rate": 9.054951373226484e-07,
|
| 1954 |
+
"loss": 0.0002,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 9.457627118644067,
|
| 1959 |
+
"grad_norm": 0.021497737616300583,
|
| 1960 |
+
"learning_rate": 8.023769889830002e-07,
|
| 1961 |
+
"loss": 0.0011,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 9.491525423728813,
|
| 1966 |
+
"grad_norm": 0.031376853585243225,
|
| 1967 |
+
"learning_rate": 7.054433187187071e-07,
|
| 1968 |
+
"loss": 0.0003,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 9.525423728813559,
|
| 1973 |
+
"grad_norm": 0.017223788425326347,
|
| 1974 |
+
"learning_rate": 6.147063117558471e-07,
|
| 1975 |
+
"loss": 0.0002,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 9.559322033898304,
|
| 1980 |
+
"grad_norm": 0.008591080084443092,
|
| 1981 |
+
"learning_rate": 5.301773743574712e-07,
|
| 1982 |
+
"loss": 0.0002,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 9.59322033898305,
|
| 1987 |
+
"grad_norm": 0.016840104013681412,
|
| 1988 |
+
"learning_rate": 4.518671323897938e-07,
|
| 1989 |
+
"loss": 0.0004,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 9.627118644067796,
|
| 1994 |
+
"grad_norm": 0.01281412597745657,
|
| 1995 |
+
"learning_rate": 3.7978542998643454e-07,
|
| 1996 |
+
"loss": 0.0001,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 9.661016949152543,
|
| 2001 |
+
"grad_norm": 0.01014797855168581,
|
| 2002 |
+
"learning_rate": 3.1394132831094047e-07,
|
| 2003 |
+
"loss": 0.0001,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 9.694915254237289,
|
| 2008 |
+
"grad_norm": 0.006094843149185181,
|
| 2009 |
+
"learning_rate": 2.5434310441773137e-07,
|
| 2010 |
+
"loss": 0.0002,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 9.728813559322035,
|
| 2015 |
+
"grad_norm": 0.054053328931331635,
|
| 2016 |
+
"learning_rate": 2.0099825021163144e-07,
|
| 2017 |
+
"loss": 0.0002,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 9.76271186440678,
|
| 2022 |
+
"grad_norm": 0.07471420615911484,
|
| 2023 |
+
"learning_rate": 1.5391347150607304e-07,
|
| 2024 |
+
"loss": 0.0005,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 9.796610169491526,
|
| 2029 |
+
"grad_norm": 0.1749231517314911,
|
| 2030 |
+
"learning_rate": 1.1309468718013194e-07,
|
| 2031 |
+
"loss": 0.0005,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 9.830508474576272,
|
| 2036 |
+
"grad_norm": 0.011661729775369167,
|
| 2037 |
+
"learning_rate": 7.854702843449469e-08,
|
| 2038 |
+
"loss": 0.0004,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 9.864406779661017,
|
| 2043 |
+
"grad_norm": 0.12255346029996872,
|
| 2044 |
+
"learning_rate": 5.0274838146413536e-08,
|
| 2045 |
+
"loss": 0.0009,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 9.898305084745763,
|
| 2050 |
+
"grad_norm": 0.008154436945915222,
|
| 2051 |
+
"learning_rate": 2.828167032379869e-08,
|
| 2052 |
+
"loss": 0.0002,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 9.932203389830509,
|
| 2057 |
+
"grad_norm": 0.017673006281256676,
|
| 2058 |
+
"learning_rate": 1.257028965842566e-08,
|
| 2059 |
+
"loss": 0.0001,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 9.966101694915254,
|
| 2064 |
+
"grad_norm": 0.009231282398104668,
|
| 2065 |
+
"learning_rate": 3.1426711784299233e-09,
|
| 2066 |
+
"loss": 0.0004,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 10.0,
|
| 2071 |
+
"grad_norm": 0.3303571045398712,
|
| 2072 |
+
"learning_rate": 0.0,
|
| 2073 |
+
"loss": 0.0006,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 10.0,
|
| 2078 |
+
"step": 2950,
|
| 2079 |
+
"total_flos": 3.087953656660627e+17,
|
| 2080 |
+
"train_loss": 0.0031311477128005884,
|
| 2081 |
+
"train_runtime": 2880.9611,
|
| 2082 |
+
"train_samples_per_second": 50.046,
|
| 2083 |
+
"train_steps_per_second": 1.024
|
| 2084 |
+
}
|
| 2085 |
+
],
|
| 2086 |
+
"logging_steps": 10,
|
| 2087 |
+
"max_steps": 2950,
|
| 2088 |
+
"num_input_tokens_seen": 0,
|
| 2089 |
+
"num_train_epochs": 10,
|
| 2090 |
+
"save_steps": 10000,
|
| 2091 |
+
"stateful_callbacks": {
|
| 2092 |
+
"TrainerControl": {
|
| 2093 |
+
"args": {
|
| 2094 |
+
"should_epoch_stop": false,
|
| 2095 |
+
"should_evaluate": false,
|
| 2096 |
+
"should_log": false,
|
| 2097 |
+
"should_save": true,
|
| 2098 |
+
"should_training_stop": true
|
| 2099 |
+
},
|
| 2100 |
+
"attributes": {}
|
| 2101 |
+
}
|
| 2102 |
+
},
|
| 2103 |
+
"total_flos": 3.087953656660627e+17,
|
| 2104 |
+
"train_batch_size": 49,
|
| 2105 |
+
"trial_name": null,
|
| 2106 |
+
"trial_params": null
|
| 2107 |
+
}
|