Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2191 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2191 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.001626016260163,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3076,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.016260162601626018,
|
| 13 |
+
"grad_norm": 11.395153045654297,
|
| 14 |
+
"learning_rate": 1.2987012987012986e-05,
|
| 15 |
+
"loss": 1.6854,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.032520325203252036,
|
| 20 |
+
"grad_norm": 3.7914836406707764,
|
| 21 |
+
"learning_rate": 2.5974025974025972e-05,
|
| 22 |
+
"loss": 0.7918,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.04878048780487805,
|
| 27 |
+
"grad_norm": 2.904379367828369,
|
| 28 |
+
"learning_rate": 3.8961038961038966e-05,
|
| 29 |
+
"loss": 0.3595,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.06504065040650407,
|
| 34 |
+
"grad_norm": 1.9180514812469482,
|
| 35 |
+
"learning_rate": 5.1948051948051944e-05,
|
| 36 |
+
"loss": 0.2492,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.08130081300813008,
|
| 41 |
+
"grad_norm": 1.5458292961120605,
|
| 42 |
+
"learning_rate": 6.493506493506494e-05,
|
| 43 |
+
"loss": 0.192,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.0975609756097561,
|
| 48 |
+
"grad_norm": 2.6975796222686768,
|
| 49 |
+
"learning_rate": 7.792207792207793e-05,
|
| 50 |
+
"loss": 0.1904,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.11382113821138211,
|
| 55 |
+
"grad_norm": 1.1650338172912598,
|
| 56 |
+
"learning_rate": 9.090909090909092e-05,
|
| 57 |
+
"loss": 0.1408,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.13008130081300814,
|
| 62 |
+
"grad_norm": 1.7408723831176758,
|
| 63 |
+
"learning_rate": 0.00010389610389610389,
|
| 64 |
+
"loss": 0.142,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.14634146341463414,
|
| 69 |
+
"grad_norm": 1.0142450332641602,
|
| 70 |
+
"learning_rate": 0.00011688311688311689,
|
| 71 |
+
"loss": 0.1426,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.16260162601626016,
|
| 76 |
+
"grad_norm": 1.9548156261444092,
|
| 77 |
+
"learning_rate": 0.00012987012987012987,
|
| 78 |
+
"loss": 0.1518,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.17886178861788618,
|
| 83 |
+
"grad_norm": 1.2884893417358398,
|
| 84 |
+
"learning_rate": 0.00014285714285714287,
|
| 85 |
+
"loss": 0.1255,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.1951219512195122,
|
| 90 |
+
"grad_norm": 0.9044198393821716,
|
| 91 |
+
"learning_rate": 0.00015584415584415587,
|
| 92 |
+
"loss": 0.1339,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.21138211382113822,
|
| 97 |
+
"grad_norm": 0.9501888155937195,
|
| 98 |
+
"learning_rate": 0.00016883116883116884,
|
| 99 |
+
"loss": 0.1031,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.22764227642276422,
|
| 104 |
+
"grad_norm": 1.059137225151062,
|
| 105 |
+
"learning_rate": 0.00018181818181818183,
|
| 106 |
+
"loss": 0.1248,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.24390243902439024,
|
| 111 |
+
"grad_norm": 1.6067619323730469,
|
| 112 |
+
"learning_rate": 0.0001948051948051948,
|
| 113 |
+
"loss": 0.1125,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.2601626016260163,
|
| 118 |
+
"grad_norm": 0.800540030002594,
|
| 119 |
+
"learning_rate": 0.00019999791929590824,
|
| 120 |
+
"loss": 0.1226,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.2764227642276423,
|
| 125 |
+
"grad_norm": 0.8457431793212891,
|
| 126 |
+
"learning_rate": 0.00019998520419557735,
|
| 127 |
+
"loss": 0.117,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.2926829268292683,
|
| 132 |
+
"grad_norm": 0.8381863236427307,
|
| 133 |
+
"learning_rate": 0.00019996093140964042,
|
| 134 |
+
"loss": 0.097,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.3089430894308943,
|
| 139 |
+
"grad_norm": 0.6342546939849854,
|
| 140 |
+
"learning_rate": 0.00019992510374388481,
|
| 141 |
+
"loss": 0.1145,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.3252032520325203,
|
| 146 |
+
"grad_norm": 1.3920717239379883,
|
| 147 |
+
"learning_rate": 0.0001998777253397723,
|
| 148 |
+
"loss": 0.1136,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.34146341463414637,
|
| 153 |
+
"grad_norm": 0.6389938592910767,
|
| 154 |
+
"learning_rate": 0.0001998188016739602,
|
| 155 |
+
"loss": 0.1014,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.35772357723577236,
|
| 160 |
+
"grad_norm": 1.111118197441101,
|
| 161 |
+
"learning_rate": 0.0001997483395576683,
|
| 162 |
+
"loss": 0.1186,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.37398373983739835,
|
| 167 |
+
"grad_norm": 0.8862912058830261,
|
| 168 |
+
"learning_rate": 0.00019966634713589157,
|
| 169 |
+
"loss": 0.1037,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.3902439024390244,
|
| 174 |
+
"grad_norm": 0.47783714532852173,
|
| 175 |
+
"learning_rate": 0.00019957283388645874,
|
| 176 |
+
"loss": 0.0986,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.4065040650406504,
|
| 181 |
+
"grad_norm": 0.7350404858589172,
|
| 182 |
+
"learning_rate": 0.00019946781061893647,
|
| 183 |
+
"loss": 0.0938,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.42276422764227645,
|
| 188 |
+
"grad_norm": 1.0116444826126099,
|
| 189 |
+
"learning_rate": 0.0001993512894733801,
|
| 190 |
+
"loss": 0.1017,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.43902439024390244,
|
| 195 |
+
"grad_norm": 0.6825156211853027,
|
| 196 |
+
"learning_rate": 0.0001992232839189301,
|
| 197 |
+
"loss": 0.0956,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.45528455284552843,
|
| 202 |
+
"grad_norm": 1.1011830568313599,
|
| 203 |
+
"learning_rate": 0.00019908380875225534,
|
| 204 |
+
"loss": 0.0993,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.4715447154471545,
|
| 209 |
+
"grad_norm": 0.560863196849823,
|
| 210 |
+
"learning_rate": 0.00019893288009584256,
|
| 211 |
+
"loss": 0.0784,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.4878048780487805,
|
| 216 |
+
"grad_norm": 0.8265031576156616,
|
| 217 |
+
"learning_rate": 0.00019877051539613264,
|
| 218 |
+
"loss": 0.0931,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.5040650406504065,
|
| 223 |
+
"grad_norm": 0.5396478772163391,
|
| 224 |
+
"learning_rate": 0.0001985967334215041,
|
| 225 |
+
"loss": 0.0804,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5203252032520326,
|
| 230 |
+
"grad_norm": 0.5586859583854675,
|
| 231 |
+
"learning_rate": 0.0001984115542601034,
|
| 232 |
+
"loss": 0.0889,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.5365853658536586,
|
| 237 |
+
"grad_norm": 0.9515529870986938,
|
| 238 |
+
"learning_rate": 0.0001982149993175229,
|
| 239 |
+
"loss": 0.0915,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.5528455284552846,
|
| 244 |
+
"grad_norm": 0.9865338206291199,
|
| 245 |
+
"learning_rate": 0.00019800709131432667,
|
| 246 |
+
"loss": 0.0765,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.5691056910569106,
|
| 251 |
+
"grad_norm": 0.46598589420318604,
|
| 252 |
+
"learning_rate": 0.00019778785428342386,
|
| 253 |
+
"loss": 0.0844,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.5853658536585366,
|
| 258 |
+
"grad_norm": 1.2223691940307617,
|
| 259 |
+
"learning_rate": 0.00019755731356729093,
|
| 260 |
+
"loss": 0.0797,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.6016260162601627,
|
| 265 |
+
"grad_norm": 0.568372905254364,
|
| 266 |
+
"learning_rate": 0.00019731549581504193,
|
| 267 |
+
"loss": 0.0713,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.6178861788617886,
|
| 272 |
+
"grad_norm": 0.9440287947654724,
|
| 273 |
+
"learning_rate": 0.00019706242897934824,
|
| 274 |
+
"loss": 0.0859,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.6341463414634146,
|
| 279 |
+
"grad_norm": 0.9875515699386597,
|
| 280 |
+
"learning_rate": 0.00019679814231320734,
|
| 281 |
+
"loss": 0.0959,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.6504065040650406,
|
| 286 |
+
"grad_norm": 1.18289315700531,
|
| 287 |
+
"learning_rate": 0.00019652266636656132,
|
| 288 |
+
"loss": 0.0794,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.6666666666666666,
|
| 293 |
+
"grad_norm": 0.8284242749214172,
|
| 294 |
+
"learning_rate": 0.00019623603298276544,
|
| 295 |
+
"loss": 0.0807,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.6829268292682927,
|
| 300 |
+
"grad_norm": 0.5042060017585754,
|
| 301 |
+
"learning_rate": 0.00019593827529490741,
|
| 302 |
+
"loss": 0.0798,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.6991869918699187,
|
| 307 |
+
"grad_norm": 0.4532330632209778,
|
| 308 |
+
"learning_rate": 0.00019562942772197723,
|
| 309 |
+
"loss": 0.0704,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.7154471544715447,
|
| 314 |
+
"grad_norm": 0.6993445754051208,
|
| 315 |
+
"learning_rate": 0.00019530952596488857,
|
| 316 |
+
"loss": 0.0829,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.7317073170731707,
|
| 321 |
+
"grad_norm": 0.7966378927230835,
|
| 322 |
+
"learning_rate": 0.00019497860700235206,
|
| 323 |
+
"loss": 0.0716,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.7479674796747967,
|
| 328 |
+
"grad_norm": 0.6459921002388,
|
| 329 |
+
"learning_rate": 0.00019463670908660075,
|
| 330 |
+
"loss": 0.0695,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.7642276422764228,
|
| 335 |
+
"grad_norm": 0.53463214635849,
|
| 336 |
+
"learning_rate": 0.00019428387173896837,
|
| 337 |
+
"loss": 0.071,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.7804878048780488,
|
| 342 |
+
"grad_norm": 0.5773355960845947,
|
| 343 |
+
"learning_rate": 0.0001939201357453208,
|
| 344 |
+
"loss": 0.0619,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.7967479674796748,
|
| 349 |
+
"grad_norm": 0.5577520728111267,
|
| 350 |
+
"learning_rate": 0.00019354554315134166,
|
| 351 |
+
"loss": 0.0643,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.8130081300813008,
|
| 356 |
+
"grad_norm": 0.6475623846054077,
|
| 357 |
+
"learning_rate": 0.0001931601372576719,
|
| 358 |
+
"loss": 0.06,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.8292682926829268,
|
| 363 |
+
"grad_norm": 0.6802220344543457,
|
| 364 |
+
"learning_rate": 0.00019276396261490465,
|
| 365 |
+
"loss": 0.0617,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.8455284552845529,
|
| 370 |
+
"grad_norm": 0.41168370842933655,
|
| 371 |
+
"learning_rate": 0.0001923570650184354,
|
| 372 |
+
"loss": 0.0529,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.8617886178861789,
|
| 377 |
+
"grad_norm": 0.5927830338478088,
|
| 378 |
+
"learning_rate": 0.00019193949150316826,
|
| 379 |
+
"loss": 0.0747,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.8780487804878049,
|
| 384 |
+
"grad_norm": 0.45662933588027954,
|
| 385 |
+
"learning_rate": 0.00019151129033807908,
|
| 386 |
+
"loss": 0.0644,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.8943089430894309,
|
| 391 |
+
"grad_norm": 0.5423726439476013,
|
| 392 |
+
"learning_rate": 0.0001910725110206358,
|
| 393 |
+
"loss": 0.0541,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.9105691056910569,
|
| 398 |
+
"grad_norm": 0.4179861843585968,
|
| 399 |
+
"learning_rate": 0.00019062320427107698,
|
| 400 |
+
"loss": 0.0558,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.926829268292683,
|
| 405 |
+
"grad_norm": 0.9079977869987488,
|
| 406 |
+
"learning_rate": 0.0001901634220265486,
|
| 407 |
+
"loss": 0.0749,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.943089430894309,
|
| 412 |
+
"grad_norm": 0.7620092034339905,
|
| 413 |
+
"learning_rate": 0.00018969321743510066,
|
| 414 |
+
"loss": 0.0692,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.959349593495935,
|
| 419 |
+
"grad_norm": 0.6436122059822083,
|
| 420 |
+
"learning_rate": 0.00018921264484954344,
|
| 421 |
+
"loss": 0.0571,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.975609756097561,
|
| 426 |
+
"grad_norm": 1.0210789442062378,
|
| 427 |
+
"learning_rate": 0.00018872175982116482,
|
| 428 |
+
"loss": 0.0646,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.991869918699187,
|
| 433 |
+
"grad_norm": 0.7681993246078491,
|
| 434 |
+
"learning_rate": 0.00018822061909330864,
|
| 435 |
+
"loss": 0.0716,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.008130081300813,
|
| 440 |
+
"grad_norm": 0.6054890751838684,
|
| 441 |
+
"learning_rate": 0.00018770928059481574,
|
| 442 |
+
"loss": 0.0656,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.024390243902439,
|
| 447 |
+
"grad_norm": 0.5673167705535889,
|
| 448 |
+
"learning_rate": 0.00018718780343332755,
|
| 449 |
+
"loss": 0.059,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.040650406504065,
|
| 454 |
+
"grad_norm": 0.5374870896339417,
|
| 455 |
+
"learning_rate": 0.00018665624788845387,
|
| 456 |
+
"loss": 0.056,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.056910569105691,
|
| 461 |
+
"grad_norm": 0.544373095035553,
|
| 462 |
+
"learning_rate": 0.0001861146754048045,
|
| 463 |
+
"loss": 0.0646,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.0731707317073171,
|
| 468 |
+
"grad_norm": 0.4903548061847687,
|
| 469 |
+
"learning_rate": 0.00018556314858488707,
|
| 470 |
+
"loss": 0.0618,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.089430894308943,
|
| 475 |
+
"grad_norm": 0.4946286082267761,
|
| 476 |
+
"learning_rate": 0.0001850017311818702,
|
| 477 |
+
"loss": 0.0484,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.1056910569105691,
|
| 482 |
+
"grad_norm": 0.46937233209609985,
|
| 483 |
+
"learning_rate": 0.00018443048809221424,
|
| 484 |
+
"loss": 0.0651,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.1219512195121952,
|
| 489 |
+
"grad_norm": 0.6026718020439148,
|
| 490 |
+
"learning_rate": 0.0001838494853481695,
|
| 491 |
+
"loss": 0.0469,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.1382113821138211,
|
| 496 |
+
"grad_norm": 0.5334330201148987,
|
| 497 |
+
"learning_rate": 0.00018325879011014352,
|
| 498 |
+
"loss": 0.0527,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.1544715447154472,
|
| 503 |
+
"grad_norm": 0.6381818652153015,
|
| 504 |
+
"learning_rate": 0.00018265847065893737,
|
| 505 |
+
"loss": 0.0621,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.170731707317073,
|
| 510 |
+
"grad_norm": 0.9406611919403076,
|
| 511 |
+
"learning_rate": 0.00018204859638785328,
|
| 512 |
+
"loss": 0.055,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.1869918699186992,
|
| 517 |
+
"grad_norm": 0.4635262191295624,
|
| 518 |
+
"learning_rate": 0.0001814292377946727,
|
| 519 |
+
"loss": 0.0594,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.203252032520325,
|
| 524 |
+
"grad_norm": 0.5755775570869446,
|
| 525 |
+
"learning_rate": 0.00018080046647350756,
|
| 526 |
+
"loss": 0.0497,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.2195121951219512,
|
| 531 |
+
"grad_norm": 0.4675120413303375,
|
| 532 |
+
"learning_rate": 0.00018016235510652425,
|
| 533 |
+
"loss": 0.0498,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.2357723577235773,
|
| 538 |
+
"grad_norm": 0.6721792817115784,
|
| 539 |
+
"learning_rate": 0.0001795149774555421,
|
| 540 |
+
"loss": 0.0562,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.2520325203252032,
|
| 545 |
+
"grad_norm": 0.416792631149292,
|
| 546 |
+
"learning_rate": 0.00017885840835350674,
|
| 547 |
+
"loss": 0.051,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.2682926829268293,
|
| 552 |
+
"grad_norm": 0.2711058259010315,
|
| 553 |
+
"learning_rate": 0.00017819272369584016,
|
| 554 |
+
"loss": 0.0549,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.2845528455284554,
|
| 559 |
+
"grad_norm": 0.6742275357246399,
|
| 560 |
+
"learning_rate": 0.00017751800043166744,
|
| 561 |
+
"loss": 0.0456,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.3008130081300813,
|
| 566 |
+
"grad_norm": 0.43051138520240784,
|
| 567 |
+
"learning_rate": 0.000176834316554922,
|
| 568 |
+
"loss": 0.0488,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.3170731707317074,
|
| 573 |
+
"grad_norm": 0.7541930079460144,
|
| 574 |
+
"learning_rate": 0.00017614175109532997,
|
| 575 |
+
"loss": 0.0531,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.3333333333333333,
|
| 580 |
+
"grad_norm": 0.46846646070480347,
|
| 581 |
+
"learning_rate": 0.00017544038410927476,
|
| 582 |
+
"loss": 0.0632,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.3495934959349594,
|
| 587 |
+
"grad_norm": 0.5363613963127136,
|
| 588 |
+
"learning_rate": 0.00017473029667054298,
|
| 589 |
+
"loss": 0.0546,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.3658536585365852,
|
| 594 |
+
"grad_norm": 0.5677704811096191,
|
| 595 |
+
"learning_rate": 0.00017401157086095317,
|
| 596 |
+
"loss": 0.0536,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.3821138211382114,
|
| 601 |
+
"grad_norm": 0.4499371647834778,
|
| 602 |
+
"learning_rate": 0.00017328428976086702,
|
| 603 |
+
"loss": 0.0549,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.3983739837398375,
|
| 608 |
+
"grad_norm": 0.42111602425575256,
|
| 609 |
+
"learning_rate": 0.00017254853743958642,
|
| 610 |
+
"loss": 0.0576,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.4146341463414633,
|
| 615 |
+
"grad_norm": 0.602150022983551,
|
| 616 |
+
"learning_rate": 0.00017180439894563497,
|
| 617 |
+
"loss": 0.0522,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.4308943089430894,
|
| 622 |
+
"grad_norm": 0.5597222447395325,
|
| 623 |
+
"learning_rate": 0.00017105196029692743,
|
| 624 |
+
"loss": 0.0526,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.4471544715447155,
|
| 629 |
+
"grad_norm": 0.3994801342487335,
|
| 630 |
+
"learning_rate": 0.00017029130847082615,
|
| 631 |
+
"loss": 0.0451,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.4634146341463414,
|
| 636 |
+
"grad_norm": 0.36083677411079407,
|
| 637 |
+
"learning_rate": 0.00016952253139408723,
|
| 638 |
+
"loss": 0.0467,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.4796747967479675,
|
| 643 |
+
"grad_norm": 0.4908004105091095,
|
| 644 |
+
"learning_rate": 0.00016874571793269665,
|
| 645 |
+
"loss": 0.0519,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.4959349593495934,
|
| 650 |
+
"grad_norm": 0.3880288600921631,
|
| 651 |
+
"learning_rate": 0.0001679609578815979,
|
| 652 |
+
"loss": 0.0603,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.5121951219512195,
|
| 657 |
+
"grad_norm": 0.30312085151672363,
|
| 658 |
+
"learning_rate": 0.00016716834195431223,
|
| 659 |
+
"loss": 0.0651,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.5284552845528454,
|
| 664 |
+
"grad_norm": 0.5956284999847412,
|
| 665 |
+
"learning_rate": 0.00016636796177245278,
|
| 666 |
+
"loss": 0.0431,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.5447154471544715,
|
| 671 |
+
"grad_norm": 0.4647449553012848,
|
| 672 |
+
"learning_rate": 0.0001655599098551335,
|
| 673 |
+
"loss": 0.0455,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.5609756097560976,
|
| 678 |
+
"grad_norm": 0.3006104528903961,
|
| 679 |
+
"learning_rate": 0.00016474427960827472,
|
| 680 |
+
"loss": 0.0575,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.5772357723577235,
|
| 685 |
+
"grad_norm": 0.37439459562301636,
|
| 686 |
+
"learning_rate": 0.0001639211653138059,
|
| 687 |
+
"loss": 0.05,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.5934959349593496,
|
| 692 |
+
"grad_norm": 0.40915167331695557,
|
| 693 |
+
"learning_rate": 0.00016309066211876708,
|
| 694 |
+
"loss": 0.0439,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.6097560975609757,
|
| 699 |
+
"grad_norm": 0.4699673652648926,
|
| 700 |
+
"learning_rate": 0.00016225286602431063,
|
| 701 |
+
"loss": 0.0475,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.6260162601626016,
|
| 706 |
+
"grad_norm": 0.5752647519111633,
|
| 707 |
+
"learning_rate": 0.00016140787387460405,
|
| 708 |
+
"loss": 0.0501,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.6422764227642277,
|
| 713 |
+
"grad_norm": 0.294344037771225,
|
| 714 |
+
"learning_rate": 0.0001605557833456354,
|
| 715 |
+
"loss": 0.0659,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.6585365853658538,
|
| 720 |
+
"grad_norm": 0.4021753668785095,
|
| 721 |
+
"learning_rate": 0.0001596966929339224,
|
| 722 |
+
"loss": 0.0401,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.6747967479674797,
|
| 727 |
+
"grad_norm": 0.5346744656562805,
|
| 728 |
+
"learning_rate": 0.00015883070194512694,
|
| 729 |
+
"loss": 0.0428,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.6910569105691056,
|
| 734 |
+
"grad_norm": 0.5238606929779053,
|
| 735 |
+
"learning_rate": 0.0001579579104825761,
|
| 736 |
+
"loss": 0.0402,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.7073170731707317,
|
| 741 |
+
"grad_norm": 0.4844679534435272,
|
| 742 |
+
"learning_rate": 0.00015707841943569037,
|
| 743 |
+
"loss": 0.0445,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.7235772357723578,
|
| 748 |
+
"grad_norm": 0.603135347366333,
|
| 749 |
+
"learning_rate": 0.000156192330468322,
|
| 750 |
+
"loss": 0.0544,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.7398373983739837,
|
| 755 |
+
"grad_norm": 0.40627196431159973,
|
| 756 |
+
"learning_rate": 0.00015529974600700275,
|
| 757 |
+
"loss": 0.0461,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.7560975609756098,
|
| 762 |
+
"grad_norm": 0.5749149918556213,
|
| 763 |
+
"learning_rate": 0.0001544007692291044,
|
| 764 |
+
"loss": 0.0434,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.7723577235772359,
|
| 769 |
+
"grad_norm": 0.39188066124916077,
|
| 770 |
+
"learning_rate": 0.0001534955040509119,
|
| 771 |
+
"loss": 0.0414,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.7886178861788617,
|
| 776 |
+
"grad_norm": 0.22225162386894226,
|
| 777 |
+
"learning_rate": 0.00015258405511561115,
|
| 778 |
+
"loss": 0.0363,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.8048780487804879,
|
| 783 |
+
"grad_norm": 0.394105464220047,
|
| 784 |
+
"learning_rate": 0.0001516665277811932,
|
| 785 |
+
"loss": 0.0407,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.821138211382114,
|
| 790 |
+
"grad_norm": 0.2789241075515747,
|
| 791 |
+
"learning_rate": 0.00015074302810827515,
|
| 792 |
+
"loss": 0.035,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.8373983739837398,
|
| 797 |
+
"grad_norm": 0.5585935115814209,
|
| 798 |
+
"learning_rate": 0.00014981366284784058,
|
| 799 |
+
"loss": 0.0433,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.8536585365853657,
|
| 804 |
+
"grad_norm": 0.7364455461502075,
|
| 805 |
+
"learning_rate": 0.00014887853942889927,
|
| 806 |
+
"loss": 0.0371,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.8699186991869918,
|
| 811 |
+
"grad_norm": 0.46286317706108093,
|
| 812 |
+
"learning_rate": 0.0001479377659460695,
|
| 813 |
+
"loss": 0.0492,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.886178861788618,
|
| 818 |
+
"grad_norm": 0.470198392868042,
|
| 819 |
+
"learning_rate": 0.00014699145114708287,
|
| 820 |
+
"loss": 0.038,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.9024390243902438,
|
| 825 |
+
"grad_norm": 0.5939562320709229,
|
| 826 |
+
"learning_rate": 0.00014603970442021347,
|
| 827 |
+
"loss": 0.0507,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.91869918699187,
|
| 832 |
+
"grad_norm": 0.4777078926563263,
|
| 833 |
+
"learning_rate": 0.00014508263578163358,
|
| 834 |
+
"loss": 0.0443,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.934959349593496,
|
| 839 |
+
"grad_norm": 0.4176855981349945,
|
| 840 |
+
"learning_rate": 0.00014412035586269624,
|
| 841 |
+
"loss": 0.0463,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.951219512195122,
|
| 846 |
+
"grad_norm": 0.493206262588501,
|
| 847 |
+
"learning_rate": 0.0001431529758971471,
|
| 848 |
+
"loss": 0.0557,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.967479674796748,
|
| 853 |
+
"grad_norm": 0.44188496470451355,
|
| 854 |
+
"learning_rate": 0.00014218060770826636,
|
| 855 |
+
"loss": 0.0561,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.9837398373983741,
|
| 860 |
+
"grad_norm": 0.5256350636482239,
|
| 861 |
+
"learning_rate": 0.00014120336369594265,
|
| 862 |
+
"loss": 0.0504,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 2.0,
|
| 867 |
+
"grad_norm": 0.2594618499279022,
|
| 868 |
+
"learning_rate": 0.0001402213568236804,
|
| 869 |
+
"loss": 0.0403,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 2.016260162601626,
|
| 874 |
+
"grad_norm": 0.36164024472236633,
|
| 875 |
+
"learning_rate": 0.00013923470060554186,
|
| 876 |
+
"loss": 0.047,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 2.032520325203252,
|
| 881 |
+
"grad_norm": 0.44333750009536743,
|
| 882 |
+
"learning_rate": 0.00013824350909302552,
|
| 883 |
+
"loss": 0.0368,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.048780487804878,
|
| 888 |
+
"grad_norm": 0.6669727563858032,
|
| 889 |
+
"learning_rate": 0.00013724789686188267,
|
| 890 |
+
"loss": 0.0305,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 2.065040650406504,
|
| 895 |
+
"grad_norm": 0.32740581035614014,
|
| 896 |
+
"learning_rate": 0.0001362479789988727,
|
| 897 |
+
"loss": 0.0347,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 2.08130081300813,
|
| 902 |
+
"grad_norm": 0.5777595043182373,
|
| 903 |
+
"learning_rate": 0.00013524387108846032,
|
| 904 |
+
"loss": 0.0476,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 2.097560975609756,
|
| 909 |
+
"grad_norm": 0.30891892313957214,
|
| 910 |
+
"learning_rate": 0.0001342356891994542,
|
| 911 |
+
"loss": 0.0437,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.113821138211382,
|
| 916 |
+
"grad_norm": 0.4573158919811249,
|
| 917 |
+
"learning_rate": 0.00013322354987159045,
|
| 918 |
+
"loss": 0.0421,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 2.130081300813008,
|
| 923 |
+
"grad_norm": 0.42825648188591003,
|
| 924 |
+
"learning_rate": 0.00013220757010206112,
|
| 925 |
+
"loss": 0.0405,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.1463414634146343,
|
| 930 |
+
"grad_norm": 0.26400211453437805,
|
| 931 |
+
"learning_rate": 0.00013118786733199014,
|
| 932 |
+
"loss": 0.0385,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 2.16260162601626,
|
| 937 |
+
"grad_norm": 0.42198774218559265,
|
| 938 |
+
"learning_rate": 0.00013016455943285786,
|
| 939 |
+
"loss": 0.0464,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 2.178861788617886,
|
| 944 |
+
"grad_norm": 0.41942107677459717,
|
| 945 |
+
"learning_rate": 0.0001291377646928757,
|
| 946 |
+
"loss": 0.0514,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 2.1951219512195124,
|
| 951 |
+
"grad_norm": 0.2822849154472351,
|
| 952 |
+
"learning_rate": 0.00012810760180331288,
|
| 953 |
+
"loss": 0.0365,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 2.2113821138211383,
|
| 958 |
+
"grad_norm": 0.41634827852249146,
|
| 959 |
+
"learning_rate": 0.00012707418984477637,
|
| 960 |
+
"loss": 0.0497,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 2.227642276422764,
|
| 965 |
+
"grad_norm": 0.3224164545536041,
|
| 966 |
+
"learning_rate": 0.0001260376482734458,
|
| 967 |
+
"loss": 0.0379,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.2439024390243905,
|
| 972 |
+
"grad_norm": 0.34206849336624146,
|
| 973 |
+
"learning_rate": 0.00012499809690726522,
|
| 974 |
+
"loss": 0.0338,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 2.2601626016260163,
|
| 979 |
+
"grad_norm": 0.364622563123703,
|
| 980 |
+
"learning_rate": 0.00012395565591209273,
|
| 981 |
+
"loss": 0.0422,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 2.2764227642276422,
|
| 986 |
+
"grad_norm": 0.28372102975845337,
|
| 987 |
+
"learning_rate": 0.00012291044578781015,
|
| 988 |
+
"loss": 0.0369,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.292682926829268,
|
| 993 |
+
"grad_norm": 0.6621326804161072,
|
| 994 |
+
"learning_rate": 0.00012186258735439379,
|
| 995 |
+
"loss": 0.0547,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 2.3089430894308944,
|
| 1000 |
+
"grad_norm": 0.3451598286628723,
|
| 1001 |
+
"learning_rate": 0.00012081220173794863,
|
| 1002 |
+
"loss": 0.0411,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 2.3252032520325203,
|
| 1007 |
+
"grad_norm": 0.36769524216651917,
|
| 1008 |
+
"learning_rate": 0.00011975941035670664,
|
| 1009 |
+
"loss": 0.0428,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.341463414634146,
|
| 1014 |
+
"grad_norm": 0.38125452399253845,
|
| 1015 |
+
"learning_rate": 0.00011870433490699166,
|
| 1016 |
+
"loss": 0.0363,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 2.3577235772357725,
|
| 1021 |
+
"grad_norm": 0.44016361236572266,
|
| 1022 |
+
"learning_rate": 0.00011764709734915218,
|
| 1023 |
+
"loss": 0.038,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 2.3739837398373984,
|
| 1028 |
+
"grad_norm": 0.2804912328720093,
|
| 1029 |
+
"learning_rate": 0.0001165878198934632,
|
| 1030 |
+
"loss": 0.0385,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 2.3902439024390243,
|
| 1035 |
+
"grad_norm": 0.31978878378868103,
|
| 1036 |
+
"learning_rate": 0.00011552662498599958,
|
| 1037 |
+
"loss": 0.0269,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 2.40650406504065,
|
| 1042 |
+
"grad_norm": 0.2746686339378357,
|
| 1043 |
+
"learning_rate": 0.00011446363529448209,
|
| 1044 |
+
"loss": 0.0365,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 2.4227642276422765,
|
| 1049 |
+
"grad_norm": 0.35929563641548157,
|
| 1050 |
+
"learning_rate": 0.00011339897369409774,
|
| 1051 |
+
"loss": 0.0399,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.4390243902439024,
|
| 1056 |
+
"grad_norm": 0.25356271862983704,
|
| 1057 |
+
"learning_rate": 0.00011233276325329596,
|
| 1058 |
+
"loss": 0.0381,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 2.4552845528455283,
|
| 1063 |
+
"grad_norm": 0.27866753935813904,
|
| 1064 |
+
"learning_rate": 0.00011126512721956289,
|
| 1065 |
+
"loss": 0.0273,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 2.4715447154471546,
|
| 1070 |
+
"grad_norm": 0.3891955018043518,
|
| 1071 |
+
"learning_rate": 0.00011019618900517455,
|
| 1072 |
+
"loss": 0.0367,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 2.4878048780487805,
|
| 1077 |
+
"grad_norm": 0.26467326283454895,
|
| 1078 |
+
"learning_rate": 0.0001091260721729312,
|
| 1079 |
+
"loss": 0.0311,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 2.5040650406504064,
|
| 1084 |
+
"grad_norm": 0.21822647750377655,
|
| 1085 |
+
"learning_rate": 0.0001080549004218742,
|
| 1086 |
+
"loss": 0.0296,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 2.5203252032520327,
|
| 1091 |
+
"grad_norm": 0.3661019206047058,
|
| 1092 |
+
"learning_rate": 0.00010698279757298715,
|
| 1093 |
+
"loss": 0.0395,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.5365853658536586,
|
| 1098 |
+
"grad_norm": 0.15822070837020874,
|
| 1099 |
+
"learning_rate": 0.00010590988755488292,
|
| 1100 |
+
"loss": 0.0265,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 2.5528455284552845,
|
| 1105 |
+
"grad_norm": 0.25284048914909363,
|
| 1106 |
+
"learning_rate": 0.00010483629438947826,
|
| 1107 |
+
"loss": 0.0307,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 2.569105691056911,
|
| 1112 |
+
"grad_norm": 0.3222500681877136,
|
| 1113 |
+
"learning_rate": 0.00010376214217765772,
|
| 1114 |
+
"loss": 0.0361,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 2.5853658536585367,
|
| 1119 |
+
"grad_norm": 0.3294614255428314,
|
| 1120 |
+
"learning_rate": 0.00010268755508492804,
|
| 1121 |
+
"loss": 0.0354,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 2.6016260162601625,
|
| 1126 |
+
"grad_norm": 0.4046204388141632,
|
| 1127 |
+
"learning_rate": 0.0001016126573270658,
|
| 1128 |
+
"loss": 0.0297,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 2.617886178861789,
|
| 1133 |
+
"grad_norm": 0.37592411041259766,
|
| 1134 |
+
"learning_rate": 0.0001005375731557584,
|
| 1135 |
+
"loss": 0.0326,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.6341463414634148,
|
| 1140 |
+
"grad_norm": 0.4793863594532013,
|
| 1141 |
+
"learning_rate": 9.946242684424162e-05,
|
| 1142 |
+
"loss": 0.0315,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 2.6504065040650406,
|
| 1147 |
+
"grad_norm": 0.31578221917152405,
|
| 1148 |
+
"learning_rate": 9.838734267293421e-05,
|
| 1149 |
+
"loss": 0.0316,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 2.6666666666666665,
|
| 1154 |
+
"grad_norm": 0.4366059899330139,
|
| 1155 |
+
"learning_rate": 9.731244491507197e-05,
|
| 1156 |
+
"loss": 0.0286,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 2.682926829268293,
|
| 1161 |
+
"grad_norm": 0.3444521725177765,
|
| 1162 |
+
"learning_rate": 9.623785782234234e-05,
|
| 1163 |
+
"loss": 0.0278,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 2.6991869918699187,
|
| 1168 |
+
"grad_norm": 0.5460310578346252,
|
| 1169 |
+
"learning_rate": 9.516370561052174e-05,
|
| 1170 |
+
"loss": 0.0444,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 2.7154471544715446,
|
| 1175 |
+
"grad_norm": 0.360115110874176,
|
| 1176 |
+
"learning_rate": 9.409011244511712e-05,
|
| 1177 |
+
"loss": 0.0288,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 2.7317073170731705,
|
| 1182 |
+
"grad_norm": 0.3582008183002472,
|
| 1183 |
+
"learning_rate": 9.30172024270129e-05,
|
| 1184 |
+
"loss": 0.0311,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 2.747967479674797,
|
| 1189 |
+
"grad_norm": 0.4152023494243622,
|
| 1190 |
+
"learning_rate": 9.194509957812582e-05,
|
| 1191 |
+
"loss": 0.0463,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 2.7642276422764227,
|
| 1196 |
+
"grad_norm": 0.2088373750448227,
|
| 1197 |
+
"learning_rate": 9.087392782706883e-05,
|
| 1198 |
+
"loss": 0.0331,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 2.7804878048780486,
|
| 1203 |
+
"grad_norm": 0.5220609307289124,
|
| 1204 |
+
"learning_rate": 8.980381099482546e-05,
|
| 1205 |
+
"loss": 0.0311,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 2.796747967479675,
|
| 1210 |
+
"grad_norm": 0.49769702553749084,
|
| 1211 |
+
"learning_rate": 8.873487278043712e-05,
|
| 1212 |
+
"loss": 0.0336,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 2.813008130081301,
|
| 1217 |
+
"grad_norm": 0.3503866195678711,
|
| 1218 |
+
"learning_rate": 8.766723674670407e-05,
|
| 1219 |
+
"loss": 0.0272,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 2.8292682926829267,
|
| 1224 |
+
"grad_norm": 0.24078375101089478,
|
| 1225 |
+
"learning_rate": 8.660102630590227e-05,
|
| 1226 |
+
"loss": 0.0365,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 2.845528455284553,
|
| 1231 |
+
"grad_norm": 0.27272090315818787,
|
| 1232 |
+
"learning_rate": 8.553636470551792e-05,
|
| 1233 |
+
"loss": 0.0252,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 2.861788617886179,
|
| 1238 |
+
"grad_norm": 0.25519323348999023,
|
| 1239 |
+
"learning_rate": 8.447337501400047e-05,
|
| 1240 |
+
"loss": 0.0286,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 2.8780487804878048,
|
| 1245 |
+
"grad_norm": 0.27832046151161194,
|
| 1246 |
+
"learning_rate": 8.341218010653684e-05,
|
| 1247 |
+
"loss": 0.0306,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 2.894308943089431,
|
| 1252 |
+
"grad_norm": 0.3626549243927002,
|
| 1253 |
+
"learning_rate": 8.235290265084785e-05,
|
| 1254 |
+
"loss": 0.0266,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 2.910569105691057,
|
| 1259 |
+
"grad_norm": 0.2994539737701416,
|
| 1260 |
+
"learning_rate": 8.129566509300835e-05,
|
| 1261 |
+
"loss": 0.0338,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.926829268292683,
|
| 1266 |
+
"grad_norm": 0.32438531517982483,
|
| 1267 |
+
"learning_rate": 8.02405896432934e-05,
|
| 1268 |
+
"loss": 0.0308,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 2.943089430894309,
|
| 1273 |
+
"grad_norm": 0.4146476686000824,
|
| 1274 |
+
"learning_rate": 7.91877982620514e-05,
|
| 1275 |
+
"loss": 0.0258,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 2.959349593495935,
|
| 1280 |
+
"grad_norm": 0.3459465503692627,
|
| 1281 |
+
"learning_rate": 7.813741264560622e-05,
|
| 1282 |
+
"loss": 0.025,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 2.975609756097561,
|
| 1287 |
+
"grad_norm": 0.3172384202480316,
|
| 1288 |
+
"learning_rate": 7.708955421218986e-05,
|
| 1289 |
+
"loss": 0.0327,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 2.991869918699187,
|
| 1294 |
+
"grad_norm": 0.5268030166625977,
|
| 1295 |
+
"learning_rate": 7.604434408790729e-05,
|
| 1296 |
+
"loss": 0.033,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 3.008130081300813,
|
| 1301 |
+
"grad_norm": 0.2980897128582001,
|
| 1302 |
+
"learning_rate": 7.50019030927348e-05,
|
| 1303 |
+
"loss": 0.0307,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 3.024390243902439,
|
| 1308 |
+
"grad_norm": 0.3045070767402649,
|
| 1309 |
+
"learning_rate": 7.396235172655422e-05,
|
| 1310 |
+
"loss": 0.025,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 3.040650406504065,
|
| 1315 |
+
"grad_norm": 0.2718939185142517,
|
| 1316 |
+
"learning_rate": 7.292581015522363e-05,
|
| 1317 |
+
"loss": 0.0229,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 3.0569105691056913,
|
| 1322 |
+
"grad_norm": 0.5520108938217163,
|
| 1323 |
+
"learning_rate": 7.189239819668711e-05,
|
| 1324 |
+
"loss": 0.0305,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 3.073170731707317,
|
| 1329 |
+
"grad_norm": 0.2900155186653137,
|
| 1330 |
+
"learning_rate": 7.086223530712433e-05,
|
| 1331 |
+
"loss": 0.0301,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 3.089430894308943,
|
| 1336 |
+
"grad_norm": 0.31466805934906006,
|
| 1337 |
+
"learning_rate": 6.983544056714215e-05,
|
| 1338 |
+
"loss": 0.0343,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 3.105691056910569,
|
| 1343 |
+
"grad_norm": 0.4351949989795685,
|
| 1344 |
+
"learning_rate": 6.881213266800988e-05,
|
| 1345 |
+
"loss": 0.0287,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 3.1219512195121952,
|
| 1350 |
+
"grad_norm": 0.29242777824401855,
|
| 1351 |
+
"learning_rate": 6.779242989793892e-05,
|
| 1352 |
+
"loss": 0.0214,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 3.138211382113821,
|
| 1357 |
+
"grad_norm": 0.27103492617607117,
|
| 1358 |
+
"learning_rate": 6.677645012840956e-05,
|
| 1359 |
+
"loss": 0.0262,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 3.154471544715447,
|
| 1364 |
+
"grad_norm": 0.46795374155044556,
|
| 1365 |
+
"learning_rate": 6.576431080054581e-05,
|
| 1366 |
+
"loss": 0.0264,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 3.1707317073170733,
|
| 1371 |
+
"grad_norm": 0.2594567537307739,
|
| 1372 |
+
"learning_rate": 6.475612891153969e-05,
|
| 1373 |
+
"loss": 0.0318,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 3.186991869918699,
|
| 1378 |
+
"grad_norm": 0.2888941168785095,
|
| 1379 |
+
"learning_rate": 6.375202100112729e-05,
|
| 1380 |
+
"loss": 0.0318,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 3.203252032520325,
|
| 1385 |
+
"grad_norm": 0.3916606307029724,
|
| 1386 |
+
"learning_rate": 6.275210313811739e-05,
|
| 1387 |
+
"loss": 0.0293,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 3.2195121951219514,
|
| 1392 |
+
"grad_norm": 0.22073742747306824,
|
| 1393 |
+
"learning_rate": 6.175649090697448e-05,
|
| 1394 |
+
"loss": 0.0321,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 3.2357723577235773,
|
| 1399 |
+
"grad_norm": 0.3153645694255829,
|
| 1400 |
+
"learning_rate": 6.0765299394458185e-05,
|
| 1401 |
+
"loss": 0.0265,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 3.252032520325203,
|
| 1406 |
+
"grad_norm": 0.5098276138305664,
|
| 1407 |
+
"learning_rate": 5.977864317631965e-05,
|
| 1408 |
+
"loss": 0.0255,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 3.2682926829268295,
|
| 1413 |
+
"grad_norm": 0.2877456843852997,
|
| 1414 |
+
"learning_rate": 5.879663630405736e-05,
|
| 1415 |
+
"loss": 0.0296,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 3.2845528455284554,
|
| 1420 |
+
"grad_norm": 0.1817747801542282,
|
| 1421 |
+
"learning_rate": 5.781939229173365e-05,
|
| 1422 |
+
"loss": 0.0225,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 3.3008130081300813,
|
| 1427 |
+
"grad_norm": 0.31750866770744324,
|
| 1428 |
+
"learning_rate": 5.684702410285292e-05,
|
| 1429 |
+
"loss": 0.0237,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 3.317073170731707,
|
| 1434 |
+
"grad_norm": 0.2775154411792755,
|
| 1435 |
+
"learning_rate": 5.5879644137303797e-05,
|
| 1436 |
+
"loss": 0.0241,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 3.3333333333333335,
|
| 1441 |
+
"grad_norm": 0.30817779898643494,
|
| 1442 |
+
"learning_rate": 5.491736421836646e-05,
|
| 1443 |
+
"loss": 0.0217,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 3.3495934959349594,
|
| 1448 |
+
"grad_norm": 0.2524275481700897,
|
| 1449 |
+
"learning_rate": 5.396029557978657e-05,
|
| 1450 |
+
"loss": 0.0278,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 3.3658536585365852,
|
| 1455 |
+
"grad_norm": 0.21641696989536285,
|
| 1456 |
+
"learning_rate": 5.3008548852917194e-05,
|
| 1457 |
+
"loss": 0.0266,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 3.3821138211382116,
|
| 1462 |
+
"grad_norm": 0.32717615365982056,
|
| 1463 |
+
"learning_rate": 5.2062234053930504e-05,
|
| 1464 |
+
"loss": 0.027,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 3.3983739837398375,
|
| 1469 |
+
"grad_norm": 0.41103601455688477,
|
| 1470 |
+
"learning_rate": 5.1121460571100776e-05,
|
| 1471 |
+
"loss": 0.0186,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 3.4146341463414633,
|
| 1476 |
+
"grad_norm": 0.14661918580532074,
|
| 1477 |
+
"learning_rate": 5.018633715215948e-05,
|
| 1478 |
+
"loss": 0.0211,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 3.430894308943089,
|
| 1483 |
+
"grad_norm": 0.22353650629520416,
|
| 1484 |
+
"learning_rate": 4.9256971891724845e-05,
|
| 1485 |
+
"loss": 0.0253,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 3.4471544715447155,
|
| 1490 |
+
"grad_norm": 0.3574526906013489,
|
| 1491 |
+
"learning_rate": 4.833347221880685e-05,
|
| 1492 |
+
"loss": 0.0308,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 3.4634146341463414,
|
| 1497 |
+
"grad_norm": 0.24301108717918396,
|
| 1498 |
+
"learning_rate": 4.7415944884388904e-05,
|
| 1499 |
+
"loss": 0.0295,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 3.4796747967479673,
|
| 1504 |
+
"grad_norm": 0.28461310267448425,
|
| 1505 |
+
"learning_rate": 4.6504495949088154e-05,
|
| 1506 |
+
"loss": 0.024,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 3.4959349593495936,
|
| 1511 |
+
"grad_norm": 0.2199145406484604,
|
| 1512 |
+
"learning_rate": 4.559923077089564e-05,
|
| 1513 |
+
"loss": 0.0188,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 3.5121951219512195,
|
| 1518 |
+
"grad_norm": 0.24626000225543976,
|
| 1519 |
+
"learning_rate": 4.470025399299728e-05,
|
| 1520 |
+
"loss": 0.0295,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 3.5284552845528454,
|
| 1525 |
+
"grad_norm": 0.1908273845911026,
|
| 1526 |
+
"learning_rate": 4.3807669531678054e-05,
|
| 1527 |
+
"loss": 0.0172,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 3.5447154471544717,
|
| 1532 |
+
"grad_norm": 0.36190247535705566,
|
| 1533 |
+
"learning_rate": 4.292158056430966e-05,
|
| 1534 |
+
"loss": 0.0156,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 3.5609756097560976,
|
| 1539 |
+
"grad_norm": 0.33335763216018677,
|
| 1540 |
+
"learning_rate": 4.204208951742393e-05,
|
| 1541 |
+
"loss": 0.0187,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 3.5772357723577235,
|
| 1546 |
+
"grad_norm": 0.2627616226673126,
|
| 1547 |
+
"learning_rate": 4.116929805487307e-05,
|
| 1548 |
+
"loss": 0.022,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 3.59349593495935,
|
| 1553 |
+
"grad_norm": 0.33589500188827515,
|
| 1554 |
+
"learning_rate": 4.0303307066077653e-05,
|
| 1555 |
+
"loss": 0.0251,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 3.6097560975609757,
|
| 1560 |
+
"grad_norm": 0.2956457734107971,
|
| 1561 |
+
"learning_rate": 3.944421665436462e-05,
|
| 1562 |
+
"loss": 0.0214,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 3.6260162601626016,
|
| 1567 |
+
"grad_norm": 0.1858706772327423,
|
| 1568 |
+
"learning_rate": 3.859212612539597e-05,
|
| 1569 |
+
"loss": 0.0269,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 3.642276422764228,
|
| 1574 |
+
"grad_norm": 0.21992811560630798,
|
| 1575 |
+
"learning_rate": 3.7747133975689385e-05,
|
| 1576 |
+
"loss": 0.0231,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 3.658536585365854,
|
| 1581 |
+
"grad_norm": 0.2564171850681305,
|
| 1582 |
+
"learning_rate": 3.690933788123296e-05,
|
| 1583 |
+
"loss": 0.0224,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 3.6747967479674797,
|
| 1588 |
+
"grad_norm": 0.2475014477968216,
|
| 1589 |
+
"learning_rate": 3.607883468619414e-05,
|
| 1590 |
+
"loss": 0.0279,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 3.6910569105691056,
|
| 1595 |
+
"grad_norm": 0.3992771506309509,
|
| 1596 |
+
"learning_rate": 3.525572039172528e-05,
|
| 1597 |
+
"loss": 0.0294,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 3.7073170731707314,
|
| 1602 |
+
"grad_norm": 0.2070523500442505,
|
| 1603 |
+
"learning_rate": 3.4440090144866536e-05,
|
| 1604 |
+
"loss": 0.0255,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 3.7235772357723578,
|
| 1609 |
+
"grad_norm": 0.3437440097332001,
|
| 1610 |
+
"learning_rate": 3.363203822754728e-05,
|
| 1611 |
+
"loss": 0.0238,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 3.7398373983739837,
|
| 1616 |
+
"grad_norm": 0.2936142385005951,
|
| 1617 |
+
"learning_rate": 3.283165804568778e-05,
|
| 1618 |
+
"loss": 0.0222,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 3.7560975609756095,
|
| 1623 |
+
"grad_norm": 0.41532281041145325,
|
| 1624 |
+
"learning_rate": 3.203904211840213e-05,
|
| 1625 |
+
"loss": 0.0194,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 3.772357723577236,
|
| 1630 |
+
"grad_norm": 0.20870675146579742,
|
| 1631 |
+
"learning_rate": 3.1254282067303355e-05,
|
| 1632 |
+
"loss": 0.0226,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 3.7886178861788617,
|
| 1637 |
+
"grad_norm": 0.22978569567203522,
|
| 1638 |
+
"learning_rate": 3.0477468605912784e-05,
|
| 1639 |
+
"loss": 0.0245,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 3.8048780487804876,
|
| 1644 |
+
"grad_norm": 0.15536998212337494,
|
| 1645 |
+
"learning_rate": 2.970869152917387e-05,
|
| 1646 |
+
"loss": 0.0183,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 3.821138211382114,
|
| 1651 |
+
"grad_norm": 0.2645808458328247,
|
| 1652 |
+
"learning_rate": 2.894803970307257e-05,
|
| 1653 |
+
"loss": 0.019,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 3.83739837398374,
|
| 1658 |
+
"grad_norm": 0.17497260868549347,
|
| 1659 |
+
"learning_rate": 2.819560105436504e-05,
|
| 1660 |
+
"loss": 0.019,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 3.8536585365853657,
|
| 1665 |
+
"grad_norm": 0.3548007905483246,
|
| 1666 |
+
"learning_rate": 2.7451462560413633e-05,
|
| 1667 |
+
"loss": 0.0233,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 3.869918699186992,
|
| 1672 |
+
"grad_norm": 0.1929531693458557,
|
| 1673 |
+
"learning_rate": 2.6715710239132973e-05,
|
| 1674 |
+
"loss": 0.0196,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 3.886178861788618,
|
| 1679 |
+
"grad_norm": 0.2864411175251007,
|
| 1680 |
+
"learning_rate": 2.5988429139046865e-05,
|
| 1681 |
+
"loss": 0.025,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 3.902439024390244,
|
| 1686 |
+
"grad_norm": 0.11372525990009308,
|
| 1687 |
+
"learning_rate": 2.5269703329456996e-05,
|
| 1688 |
+
"loss": 0.0163,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 3.91869918699187,
|
| 1693 |
+
"grad_norm": 0.23025590181350708,
|
| 1694 |
+
"learning_rate": 2.4559615890725284e-05,
|
| 1695 |
+
"loss": 0.0199,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 3.934959349593496,
|
| 1700 |
+
"grad_norm": 0.14693354070186615,
|
| 1701 |
+
"learning_rate": 2.3858248904670056e-05,
|
| 1702 |
+
"loss": 0.0168,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 3.951219512195122,
|
| 1707 |
+
"grad_norm": 0.18080760538578033,
|
| 1708 |
+
"learning_rate": 2.316568344507799e-05,
|
| 1709 |
+
"loss": 0.0145,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 3.9674796747967482,
|
| 1714 |
+
"grad_norm": 0.2424541711807251,
|
| 1715 |
+
"learning_rate": 2.248199956833258e-05,
|
| 1716 |
+
"loss": 0.0181,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 3.983739837398374,
|
| 1721 |
+
"grad_norm": 0.24186311662197113,
|
| 1722 |
+
"learning_rate": 2.1807276304159873e-05,
|
| 1723 |
+
"loss": 0.0235,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 4.0,
|
| 1728 |
+
"grad_norm": 0.5643988847732544,
|
| 1729 |
+
"learning_rate": 2.1141591646493275e-05,
|
| 1730 |
+
"loss": 0.0222,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 4.016260162601626,
|
| 1735 |
+
"grad_norm": 0.21844126284122467,
|
| 1736 |
+
"learning_rate": 2.0485022544457933e-05,
|
| 1737 |
+
"loss": 0.0151,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 4.032520325203252,
|
| 1742 |
+
"grad_norm": 0.23713727295398712,
|
| 1743 |
+
"learning_rate": 1.983764489347574e-05,
|
| 1744 |
+
"loss": 0.0224,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 4.048780487804878,
|
| 1749 |
+
"grad_norm": 0.16830649971961975,
|
| 1750 |
+
"learning_rate": 1.9199533526492463e-05,
|
| 1751 |
+
"loss": 0.0243,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 4.065040650406504,
|
| 1756 |
+
"grad_norm": 0.31435462832450867,
|
| 1757 |
+
"learning_rate": 1.857076220532734e-05,
|
| 1758 |
+
"loss": 0.0229,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 4.08130081300813,
|
| 1763 |
+
"grad_norm": 0.4781794548034668,
|
| 1764 |
+
"learning_rate": 1.7951403612146744e-05,
|
| 1765 |
+
"loss": 0.0186,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 4.097560975609756,
|
| 1770 |
+
"grad_norm": 0.21629083156585693,
|
| 1771 |
+
"learning_rate": 1.7341529341062623e-05,
|
| 1772 |
+
"loss": 0.0168,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 4.1138211382113825,
|
| 1777 |
+
"grad_norm": 0.14135605096817017,
|
| 1778 |
+
"learning_rate": 1.6741209889856513e-05,
|
| 1779 |
+
"loss": 0.0178,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 4.130081300813008,
|
| 1784 |
+
"grad_norm": 0.1877690851688385,
|
| 1785 |
+
"learning_rate": 1.6150514651830483e-05,
|
| 1786 |
+
"loss": 0.0252,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 4.146341463414634,
|
| 1791 |
+
"grad_norm": 0.5082934498786926,
|
| 1792 |
+
"learning_rate": 1.5569511907785783e-05,
|
| 1793 |
+
"loss": 0.0164,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 4.16260162601626,
|
| 1798 |
+
"grad_norm": 0.20046326518058777,
|
| 1799 |
+
"learning_rate": 1.4998268818129802e-05,
|
| 1800 |
+
"loss": 0.0209,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 4.178861788617886,
|
| 1805 |
+
"grad_norm": 0.15702232718467712,
|
| 1806 |
+
"learning_rate": 1.443685141511294e-05,
|
| 1807 |
+
"loss": 0.0263,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 4.195121951219512,
|
| 1812 |
+
"grad_norm": 0.2763630449771881,
|
| 1813 |
+
"learning_rate": 1.3885324595195493e-05,
|
| 1814 |
+
"loss": 0.0218,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 4.211382113821138,
|
| 1819 |
+
"grad_norm": 0.1242925375699997,
|
| 1820 |
+
"learning_rate": 1.3343752111546138e-05,
|
| 1821 |
+
"loss": 0.0139,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 4.227642276422764,
|
| 1826 |
+
"grad_norm": 0.23633089661598206,
|
| 1827 |
+
"learning_rate": 1.2812196566672441e-05,
|
| 1828 |
+
"loss": 0.0192,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 4.2439024390243905,
|
| 1833 |
+
"grad_norm": 0.1781703680753708,
|
| 1834 |
+
"learning_rate": 1.2290719405184291e-05,
|
| 1835 |
+
"loss": 0.0182,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 4.260162601626016,
|
| 1840 |
+
"grad_norm": 0.20754830539226532,
|
| 1841 |
+
"learning_rate": 1.1779380906691361e-05,
|
| 1842 |
+
"loss": 0.0162,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 4.276422764227642,
|
| 1847 |
+
"grad_norm": 0.3521351218223572,
|
| 1848 |
+
"learning_rate": 1.1278240178835197e-05,
|
| 1849 |
+
"loss": 0.0233,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 4.2926829268292686,
|
| 1854 |
+
"grad_norm": 0.2133350968360901,
|
| 1855 |
+
"learning_rate": 1.0787355150456546e-05,
|
| 1856 |
+
"loss": 0.019,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 4.308943089430894,
|
| 1861 |
+
"grad_norm": 0.22535590827465057,
|
| 1862 |
+
"learning_rate": 1.0306782564899352e-05,
|
| 1863 |
+
"loss": 0.0174,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 4.32520325203252,
|
| 1868 |
+
"grad_norm": 0.361722856760025,
|
| 1869 |
+
"learning_rate": 9.83657797345141e-06,
|
| 1870 |
+
"loss": 0.016,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 4.341463414634147,
|
| 1875 |
+
"grad_norm": 0.19668163359165192,
|
| 1876 |
+
"learning_rate": 9.376795728923016e-06,
|
| 1877 |
+
"loss": 0.0239,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 4.357723577235772,
|
| 1882 |
+
"grad_norm": 0.1902504563331604,
|
| 1883 |
+
"learning_rate": 8.927488979364184e-06,
|
| 1884 |
+
"loss": 0.0148,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 4.373983739837398,
|
| 1889 |
+
"grad_norm": 0.344314306974411,
|
| 1890 |
+
"learning_rate": 8.488709661920946e-06,
|
| 1891 |
+
"loss": 0.0132,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 4.390243902439025,
|
| 1896 |
+
"grad_norm": 0.12004625797271729,
|
| 1897 |
+
"learning_rate": 8.060508496831743e-06,
|
| 1898 |
+
"loss": 0.0161,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 4.40650406504065,
|
| 1903 |
+
"grad_norm": 0.22964505851268768,
|
| 1904 |
+
"learning_rate": 7.642934981564609e-06,
|
| 1905 |
+
"loss": 0.014,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 4.4227642276422765,
|
| 1910 |
+
"grad_norm": 0.27576199173927307,
|
| 1911 |
+
"learning_rate": 7.23603738509534e-06,
|
| 1912 |
+
"loss": 0.0229,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 4.439024390243903,
|
| 1917 |
+
"grad_norm": 0.12927354872226715,
|
| 1918 |
+
"learning_rate": 6.839862742328107e-06,
|
| 1919 |
+
"loss": 0.0104,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 4.455284552845528,
|
| 1924 |
+
"grad_norm": 0.25259020924568176,
|
| 1925 |
+
"learning_rate": 6.454456848658363e-06,
|
| 1926 |
+
"loss": 0.0112,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 4.471544715447155,
|
| 1931 |
+
"grad_norm": 0.22319383919239044,
|
| 1932 |
+
"learning_rate": 6.079864254679191e-06,
|
| 1933 |
+
"loss": 0.0149,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 4.487804878048781,
|
| 1938 |
+
"grad_norm": 0.19844867289066315,
|
| 1939 |
+
"learning_rate": 5.716128261031628e-06,
|
| 1940 |
+
"loss": 0.0168,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 4.504065040650406,
|
| 1945 |
+
"grad_norm": 0.23324230313301086,
|
| 1946 |
+
"learning_rate": 5.363290913399232e-06,
|
| 1947 |
+
"loss": 0.014,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 4.520325203252033,
|
| 1952 |
+
"grad_norm": 0.22129973769187927,
|
| 1953 |
+
"learning_rate": 5.021392997647933e-06,
|
| 1954 |
+
"loss": 0.0155,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 4.536585365853659,
|
| 1959 |
+
"grad_norm": 0.4064948260784149,
|
| 1960 |
+
"learning_rate": 4.690474035111448e-06,
|
| 1961 |
+
"loss": 0.0202,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 4.5528455284552845,
|
| 1966 |
+
"grad_norm": 0.22724312543869019,
|
| 1967 |
+
"learning_rate": 4.370572278022788e-06,
|
| 1968 |
+
"loss": 0.0164,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 4.569105691056911,
|
| 1973 |
+
"grad_norm": 0.22973546385765076,
|
| 1974 |
+
"learning_rate": 4.061724705092574e-06,
|
| 1975 |
+
"loss": 0.0111,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 4.585365853658536,
|
| 1980 |
+
"grad_norm": 0.1258576661348343,
|
| 1981 |
+
"learning_rate": 3.76396701723456e-06,
|
| 1982 |
+
"loss": 0.0112,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 4.6016260162601625,
|
| 1987 |
+
"grad_norm": 0.2755131423473358,
|
| 1988 |
+
"learning_rate": 3.4773336334387064e-06,
|
| 1989 |
+
"loss": 0.0223,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 4.617886178861789,
|
| 1994 |
+
"grad_norm": 0.21121273934841156,
|
| 1995 |
+
"learning_rate": 3.2018576867926643e-06,
|
| 1996 |
+
"loss": 0.0156,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 4.634146341463414,
|
| 2001 |
+
"grad_norm": 0.0909205749630928,
|
| 2002 |
+
"learning_rate": 2.937571020651775e-06,
|
| 2003 |
+
"loss": 0.0162,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 4.650406504065041,
|
| 2008 |
+
"grad_norm": 0.19245898723602295,
|
| 2009 |
+
"learning_rate": 2.6845041849581033e-06,
|
| 2010 |
+
"loss": 0.0188,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 4.666666666666667,
|
| 2015 |
+
"grad_norm": 0.24272821843624115,
|
| 2016 |
+
"learning_rate": 2.442686432709096e-06,
|
| 2017 |
+
"loss": 0.0191,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 4.682926829268292,
|
| 2022 |
+
"grad_norm": 0.1934613585472107,
|
| 2023 |
+
"learning_rate": 2.212145716576142e-06,
|
| 2024 |
+
"loss": 0.0158,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 4.699186991869919,
|
| 2029 |
+
"grad_norm": 0.19412636756896973,
|
| 2030 |
+
"learning_rate": 1.9929086856733447e-06,
|
| 2031 |
+
"loss": 0.0164,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 4.715447154471545,
|
| 2036 |
+
"grad_norm": 0.26893916726112366,
|
| 2037 |
+
"learning_rate": 1.7850006824771004e-06,
|
| 2038 |
+
"loss": 0.0194,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 4.7317073170731705,
|
| 2043 |
+
"grad_norm": 0.20123203098773956,
|
| 2044 |
+
"learning_rate": 1.5884457398966046e-06,
|
| 2045 |
+
"loss": 0.0188,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 4.747967479674797,
|
| 2050 |
+
"grad_norm": 0.13909615576267242,
|
| 2051 |
+
"learning_rate": 1.4032665784958877e-06,
|
| 2052 |
+
"loss": 0.0162,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 4.764227642276423,
|
| 2057 |
+
"grad_norm": 0.17143380641937256,
|
| 2058 |
+
"learning_rate": 1.229484603867348e-06,
|
| 2059 |
+
"loss": 0.0165,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 4.780487804878049,
|
| 2064 |
+
"grad_norm": 0.11365949362516403,
|
| 2065 |
+
"learning_rate": 1.0671199041574542e-06,
|
| 2066 |
+
"loss": 0.0221,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 4.796747967479675,
|
| 2071 |
+
"grad_norm": 0.22491714358329773,
|
| 2072 |
+
"learning_rate": 9.161912477446688e-07,
|
| 2073 |
+
"loss": 0.0252,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 4.8130081300813,
|
| 2078 |
+
"grad_norm": 0.3459233343601227,
|
| 2079 |
+
"learning_rate": 7.76716081069917e-07,
|
| 2080 |
+
"loss": 0.0167,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 4.829268292682927,
|
| 2085 |
+
"grad_norm": 0.23336616158485413,
|
| 2086 |
+
"learning_rate": 6.487105266199333e-07,
|
| 2087 |
+
"loss": 0.0156,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 4.845528455284553,
|
| 2092 |
+
"grad_norm": 0.13937050104141235,
|
| 2093 |
+
"learning_rate": 5.321893810635525e-07,
|
| 2094 |
+
"loss": 0.0148,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 4.861788617886178,
|
| 2099 |
+
"grad_norm": 0.34142452478408813,
|
| 2100 |
+
"learning_rate": 4.2716611354127747e-07,
|
| 2101 |
+
"loss": 0.0156,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 4.878048780487805,
|
| 2106 |
+
"grad_norm": 0.1981387436389923,
|
| 2107 |
+
"learning_rate": 3.3365286410842465e-07,
|
| 2108 |
+
"loss": 0.029,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 4.894308943089431,
|
| 2113 |
+
"grad_norm": 0.12347165495157242,
|
| 2114 |
+
"learning_rate": 2.516604423317248e-07,
|
| 2115 |
+
"loss": 0.016,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 4.9105691056910565,
|
| 2120 |
+
"grad_norm": 0.17062854766845703,
|
| 2121 |
+
"learning_rate": 1.81198326039822e-07,
|
| 2122 |
+
"loss": 0.0165,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 4.926829268292683,
|
| 2127 |
+
"grad_norm": 0.16616424918174744,
|
| 2128 |
+
"learning_rate": 1.2227466022770584e-07,
|
| 2129 |
+
"loss": 0.0167,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 4.943089430894309,
|
| 2134 |
+
"grad_norm": 0.22849665582180023,
|
| 2135 |
+
"learning_rate": 7.489625611519779e-08,
|
| 2136 |
+
"loss": 0.0173,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 4.959349593495935,
|
| 2141 |
+
"grad_norm": 0.3534289598464966,
|
| 2142 |
+
"learning_rate": 3.906859035960331e-08,
|
| 2143 |
+
"loss": 0.0238,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 4.975609756097561,
|
| 2148 |
+
"grad_norm": 0.19245095551013947,
|
| 2149 |
+
"learning_rate": 1.4795804422651494e-08,
|
| 2150 |
+
"loss": 0.0134,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 4.991869918699187,
|
| 2155 |
+
"grad_norm": 0.15183484554290771,
|
| 2156 |
+
"learning_rate": 2.080704091766972e-09,
|
| 2157 |
+
"loss": 0.0118,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 5.001626016260163,
|
| 2162 |
+
"step": 3076,
|
| 2163 |
+
"total_flos": 1.078733099791872e+17,
|
| 2164 |
+
"train_loss": 0.05316672712002526,
|
| 2165 |
+
"train_runtime": 1399.9912,
|
| 2166 |
+
"train_samples_per_second": 35.155,
|
| 2167 |
+
"train_steps_per_second": 2.197
|
| 2168 |
+
}
|
| 2169 |
+
],
|
| 2170 |
+
"logging_steps": 10,
|
| 2171 |
+
"max_steps": 3076,
|
| 2172 |
+
"num_input_tokens_seen": 0,
|
| 2173 |
+
"num_train_epochs": 6,
|
| 2174 |
+
"save_steps": 500,
|
| 2175 |
+
"stateful_callbacks": {
|
| 2176 |
+
"TrainerControl": {
|
| 2177 |
+
"args": {
|
| 2178 |
+
"should_epoch_stop": false,
|
| 2179 |
+
"should_evaluate": false,
|
| 2180 |
+
"should_log": false,
|
| 2181 |
+
"should_save": true,
|
| 2182 |
+
"should_training_stop": true
|
| 2183 |
+
},
|
| 2184 |
+
"attributes": {}
|
| 2185 |
+
}
|
| 2186 |
+
},
|
| 2187 |
+
"total_flos": 1.078733099791872e+17,
|
| 2188 |
+
"train_batch_size": 16,
|
| 2189 |
+
"trial_name": null,
|
| 2190 |
+
"trial_params": null
|
| 2191 |
+
}
|