Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2406 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2406 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 33984,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.011770244821092278,
|
| 13 |
+
"grad_norm": 4.125,
|
| 14 |
+
"learning_rate": 1.9607843137254903e-05,
|
| 15 |
+
"loss": 0.6291,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.023540489642184557,
|
| 20 |
+
"grad_norm": 13.0625,
|
| 21 |
+
"learning_rate": 3.9215686274509805e-05,
|
| 22 |
+
"loss": 0.6276,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.03531073446327684,
|
| 27 |
+
"grad_norm": 4.09375,
|
| 28 |
+
"learning_rate": 5.882352941176471e-05,
|
| 29 |
+
"loss": 0.6124,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.047080979284369114,
|
| 34 |
+
"grad_norm": 12.9375,
|
| 35 |
+
"learning_rate": 7.843137254901961e-05,
|
| 36 |
+
"loss": 0.5819,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.05885122410546139,
|
| 41 |
+
"grad_norm": 2.296875,
|
| 42 |
+
"learning_rate": 9.80392156862745e-05,
|
| 43 |
+
"loss": 0.5751,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.07062146892655367,
|
| 48 |
+
"grad_norm": 8.9375,
|
| 49 |
+
"learning_rate": 0.00011764705882352942,
|
| 50 |
+
"loss": 0.5796,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.08239171374764595,
|
| 55 |
+
"grad_norm": 8.5625,
|
| 56 |
+
"learning_rate": 0.0001372549019607843,
|
| 57 |
+
"loss": 0.5595,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.09416195856873823,
|
| 62 |
+
"grad_norm": 3.5625,
|
| 63 |
+
"learning_rate": 0.00015686274509803922,
|
| 64 |
+
"loss": 0.539,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.1059322033898305,
|
| 69 |
+
"grad_norm": 8.3125,
|
| 70 |
+
"learning_rate": 0.00017647058823529413,
|
| 71 |
+
"loss": 0.5309,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.11770244821092278,
|
| 76 |
+
"grad_norm": 11.6875,
|
| 77 |
+
"learning_rate": 0.000196078431372549,
|
| 78 |
+
"loss": 0.5169,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.12947269303201506,
|
| 83 |
+
"grad_norm": 15.25,
|
| 84 |
+
"learning_rate": 0.00019999709351685367,
|
| 85 |
+
"loss": 0.5094,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.14124293785310735,
|
| 90 |
+
"grad_norm": 2.734375,
|
| 91 |
+
"learning_rate": 0.00019998528621863396,
|
| 92 |
+
"loss": 0.5228,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.1530131826741996,
|
| 97 |
+
"grad_norm": 2.59375,
|
| 98 |
+
"learning_rate": 0.00019996439752174486,
|
| 99 |
+
"loss": 0.51,
|
| 100 |
+
"step": 1300
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.1647834274952919,
|
| 104 |
+
"grad_norm": 14.25,
|
| 105 |
+
"learning_rate": 0.00019993442932345136,
|
| 106 |
+
"loss": 0.5085,
|
| 107 |
+
"step": 1400
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.1765536723163842,
|
| 111 |
+
"grad_norm": 6.15625,
|
| 112 |
+
"learning_rate": 0.0001998953843456855,
|
| 113 |
+
"loss": 0.4933,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.18832391713747645,
|
| 118 |
+
"grad_norm": 18.0,
|
| 119 |
+
"learning_rate": 0.00019984726613479922,
|
| 120 |
+
"loss": 0.4831,
|
| 121 |
+
"step": 1600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.20009416195856874,
|
| 125 |
+
"grad_norm": 5.34375,
|
| 126 |
+
"learning_rate": 0.0001997900790612421,
|
| 127 |
+
"loss": 0.4831,
|
| 128 |
+
"step": 1700
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.211864406779661,
|
| 132 |
+
"grad_norm": 19.625,
|
| 133 |
+
"learning_rate": 0.00019972382831916457,
|
| 134 |
+
"loss": 0.4857,
|
| 135 |
+
"step": 1800
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.2236346516007533,
|
| 139 |
+
"grad_norm": 6.59375,
|
| 140 |
+
"learning_rate": 0.000199648519925946,
|
| 141 |
+
"loss": 0.4772,
|
| 142 |
+
"step": 1900
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.23540489642184556,
|
| 146 |
+
"grad_norm": 5.21875,
|
| 147 |
+
"learning_rate": 0.00019956416072164826,
|
| 148 |
+
"loss": 0.4953,
|
| 149 |
+
"step": 2000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.24717514124293785,
|
| 153 |
+
"grad_norm": 16.25,
|
| 154 |
+
"learning_rate": 0.00019947075836839438,
|
| 155 |
+
"loss": 0.4906,
|
| 156 |
+
"step": 2100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.2589453860640301,
|
| 160 |
+
"grad_norm": 11.125,
|
| 161 |
+
"learning_rate": 0.00019936832134967265,
|
| 162 |
+
"loss": 0.4731,
|
| 163 |
+
"step": 2200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.2707156308851224,
|
| 167 |
+
"grad_norm": 6.40625,
|
| 168 |
+
"learning_rate": 0.00019925685896956616,
|
| 169 |
+
"loss": 0.472,
|
| 170 |
+
"step": 2300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.2824858757062147,
|
| 174 |
+
"grad_norm": 4.96875,
|
| 175 |
+
"learning_rate": 0.00019913638135190756,
|
| 176 |
+
"loss": 0.4788,
|
| 177 |
+
"step": 2400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.294256120527307,
|
| 181 |
+
"grad_norm": 3.9375,
|
| 182 |
+
"learning_rate": 0.00019900689943935971,
|
| 183 |
+
"loss": 0.4721,
|
| 184 |
+
"step": 2500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.3060263653483992,
|
| 188 |
+
"grad_norm": 8.6875,
|
| 189 |
+
"learning_rate": 0.0001988684249924217,
|
| 190 |
+
"loss": 0.469,
|
| 191 |
+
"step": 2600
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.3177966101694915,
|
| 195 |
+
"grad_norm": 10.75,
|
| 196 |
+
"learning_rate": 0.00019872097058836076,
|
| 197 |
+
"loss": 0.4496,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.3295668549905838,
|
| 202 |
+
"grad_norm": 3.0,
|
| 203 |
+
"learning_rate": 0.00019856454962006973,
|
| 204 |
+
"loss": 0.4528,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.3413370998116761,
|
| 209 |
+
"grad_norm": 20.0,
|
| 210 |
+
"learning_rate": 0.00019839917629485077,
|
| 211 |
+
"loss": 0.4763,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.3531073446327684,
|
| 216 |
+
"grad_norm": 8.6875,
|
| 217 |
+
"learning_rate": 0.0001982248656331249,
|
| 218 |
+
"loss": 0.4762,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.3648775894538606,
|
| 223 |
+
"grad_norm": 9.0625,
|
| 224 |
+
"learning_rate": 0.0001980416334670678,
|
| 225 |
+
"loss": 0.4758,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.3766478342749529,
|
| 230 |
+
"grad_norm": 9.8125,
|
| 231 |
+
"learning_rate": 0.00019784949643917164,
|
| 232 |
+
"loss": 0.4649,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.3884180790960452,
|
| 237 |
+
"grad_norm": 27.875,
|
| 238 |
+
"learning_rate": 0.0001976484720007337,
|
| 239 |
+
"loss": 0.4669,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.4001883239171375,
|
| 244 |
+
"grad_norm": 2.296875,
|
| 245 |
+
"learning_rate": 0.00019743857841027115,
|
| 246 |
+
"loss": 0.463,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.4119585687382298,
|
| 251 |
+
"grad_norm": 13.1875,
|
| 252 |
+
"learning_rate": 0.00019721983473186278,
|
| 253 |
+
"loss": 0.4717,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.423728813559322,
|
| 258 |
+
"grad_norm": 2.390625,
|
| 259 |
+
"learning_rate": 0.00019699226083341742,
|
| 260 |
+
"loss": 0.4766,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.4354990583804143,
|
| 265 |
+
"grad_norm": 7.28125,
|
| 266 |
+
"learning_rate": 0.00019675587738486936,
|
| 267 |
+
"loss": 0.471,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.4472693032015066,
|
| 272 |
+
"grad_norm": 2.109375,
|
| 273 |
+
"learning_rate": 0.00019651070585630104,
|
| 274 |
+
"loss": 0.4591,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.4590395480225989,
|
| 279 |
+
"grad_norm": 3.796875,
|
| 280 |
+
"learning_rate": 0.00019625676851599288,
|
| 281 |
+
"loss": 0.4778,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.4708097928436911,
|
| 286 |
+
"grad_norm": 12.4375,
|
| 287 |
+
"learning_rate": 0.0001959940884284008,
|
| 288 |
+
"loss": 0.4474,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.4825800376647834,
|
| 293 |
+
"grad_norm": 8.8125,
|
| 294 |
+
"learning_rate": 0.00019572268945206126,
|
| 295 |
+
"loss": 0.4669,
|
| 296 |
+
"step": 4100
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.4943502824858757,
|
| 300 |
+
"grad_norm": 2.96875,
|
| 301 |
+
"learning_rate": 0.00019544259623742428,
|
| 302 |
+
"loss": 0.4736,
|
| 303 |
+
"step": 4200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.506120527306968,
|
| 307 |
+
"grad_norm": 3.796875,
|
| 308 |
+
"learning_rate": 0.00019515383422461454,
|
| 309 |
+
"loss": 0.4709,
|
| 310 |
+
"step": 4300
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.5178907721280602,
|
| 314 |
+
"grad_norm": 11.5,
|
| 315 |
+
"learning_rate": 0.0001948564296411207,
|
| 316 |
+
"loss": 0.4612,
|
| 317 |
+
"step": 4400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.5296610169491526,
|
| 321 |
+
"grad_norm": 2.328125,
|
| 322 |
+
"learning_rate": 0.0001945504094994132,
|
| 323 |
+
"loss": 0.4586,
|
| 324 |
+
"step": 4500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.5414312617702448,
|
| 328 |
+
"grad_norm": 9.6875,
|
| 329 |
+
"learning_rate": 0.00019423580159449083,
|
| 330 |
+
"loss": 0.4488,
|
| 331 |
+
"step": 4600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.553201506591337,
|
| 335 |
+
"grad_norm": 9.25,
|
| 336 |
+
"learning_rate": 0.00019391263450135607,
|
| 337 |
+
"loss": 0.4721,
|
| 338 |
+
"step": 4700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.5649717514124294,
|
| 342 |
+
"grad_norm": 6.28125,
|
| 343 |
+
"learning_rate": 0.00019358093757241996,
|
| 344 |
+
"loss": 0.4507,
|
| 345 |
+
"step": 4800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.5767419962335216,
|
| 349 |
+
"grad_norm": 6.84375,
|
| 350 |
+
"learning_rate": 0.00019324074093483577,
|
| 351 |
+
"loss": 0.4579,
|
| 352 |
+
"step": 4900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.588512241054614,
|
| 356 |
+
"grad_norm": 5.3125,
|
| 357 |
+
"learning_rate": 0.0001928920754877629,
|
| 358 |
+
"loss": 0.4778,
|
| 359 |
+
"step": 5000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.6002824858757062,
|
| 363 |
+
"grad_norm": 1.6875,
|
| 364 |
+
"learning_rate": 0.0001925349728995602,
|
| 365 |
+
"loss": 0.4615,
|
| 366 |
+
"step": 5100
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.6120527306967984,
|
| 370 |
+
"grad_norm": 9.4375,
|
| 371 |
+
"learning_rate": 0.00019216946560490973,
|
| 372 |
+
"loss": 0.4628,
|
| 373 |
+
"step": 5200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.6238229755178908,
|
| 377 |
+
"grad_norm": 3.078125,
|
| 378 |
+
"learning_rate": 0.00019179558680187077,
|
| 379 |
+
"loss": 0.4768,
|
| 380 |
+
"step": 5300
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.635593220338983,
|
| 384 |
+
"grad_norm": 12.8125,
|
| 385 |
+
"learning_rate": 0.00019141337044886457,
|
| 386 |
+
"loss": 0.4498,
|
| 387 |
+
"step": 5400
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.6473634651600754,
|
| 391 |
+
"grad_norm": 3.46875,
|
| 392 |
+
"learning_rate": 0.0001910228512615899,
|
| 393 |
+
"loss": 0.4535,
|
| 394 |
+
"step": 5500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.6591337099811676,
|
| 398 |
+
"grad_norm": 14.5625,
|
| 399 |
+
"learning_rate": 0.00019062406470987,
|
| 400 |
+
"loss": 0.4484,
|
| 401 |
+
"step": 5600
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.6709039548022598,
|
| 405 |
+
"grad_norm": 13.0,
|
| 406 |
+
"learning_rate": 0.00019021704701443083,
|
| 407 |
+
"loss": 0.4697,
|
| 408 |
+
"step": 5700
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.6826741996233522,
|
| 412 |
+
"grad_norm": 10.25,
|
| 413 |
+
"learning_rate": 0.00018980183514361154,
|
| 414 |
+
"loss": 0.462,
|
| 415 |
+
"step": 5800
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.6944444444444444,
|
| 419 |
+
"grad_norm": 2.4375,
|
| 420 |
+
"learning_rate": 0.00018937846681000631,
|
| 421 |
+
"loss": 0.4562,
|
| 422 |
+
"step": 5900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.7062146892655368,
|
| 426 |
+
"grad_norm": 4.3125,
|
| 427 |
+
"learning_rate": 0.0001889469804670393,
|
| 428 |
+
"loss": 0.4509,
|
| 429 |
+
"step": 6000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.717984934086629,
|
| 433 |
+
"grad_norm": 14.625,
|
| 434 |
+
"learning_rate": 0.000188507415305472,
|
| 435 |
+
"loss": 0.4401,
|
| 436 |
+
"step": 6100
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.7297551789077212,
|
| 440 |
+
"grad_norm": 5.96875,
|
| 441 |
+
"learning_rate": 0.0001880598112498435,
|
| 442 |
+
"loss": 0.4609,
|
| 443 |
+
"step": 6200
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.7415254237288136,
|
| 447 |
+
"grad_norm": 12.25,
|
| 448 |
+
"learning_rate": 0.00018760420895484446,
|
| 449 |
+
"loss": 0.4496,
|
| 450 |
+
"step": 6300
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.7532956685499058,
|
| 454 |
+
"grad_norm": 3.03125,
|
| 455 |
+
"learning_rate": 0.00018714064980162435,
|
| 456 |
+
"loss": 0.4541,
|
| 457 |
+
"step": 6400
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.7650659133709982,
|
| 461 |
+
"grad_norm": 15.4375,
|
| 462 |
+
"learning_rate": 0.00018666917589403297,
|
| 463 |
+
"loss": 0.4544,
|
| 464 |
+
"step": 6500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.7768361581920904,
|
| 468 |
+
"grad_norm": 5.21875,
|
| 469 |
+
"learning_rate": 0.00018618983005479637,
|
| 470 |
+
"loss": 0.4532,
|
| 471 |
+
"step": 6600
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.7886064030131826,
|
| 475 |
+
"grad_norm": 4.96875,
|
| 476 |
+
"learning_rate": 0.00018570265582162728,
|
| 477 |
+
"loss": 0.454,
|
| 478 |
+
"step": 6700
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.800376647834275,
|
| 482 |
+
"grad_norm": 8.4375,
|
| 483 |
+
"learning_rate": 0.00018520769744327062,
|
| 484 |
+
"loss": 0.4576,
|
| 485 |
+
"step": 6800
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.8121468926553672,
|
| 489 |
+
"grad_norm": 2.640625,
|
| 490 |
+
"learning_rate": 0.00018470499987548473,
|
| 491 |
+
"loss": 0.4578,
|
| 492 |
+
"step": 6900
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.8239171374764596,
|
| 496 |
+
"grad_norm": 10.4375,
|
| 497 |
+
"learning_rate": 0.00018419460877695795,
|
| 498 |
+
"loss": 0.4544,
|
| 499 |
+
"step": 7000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.8356873822975518,
|
| 503 |
+
"grad_norm": 25.25,
|
| 504 |
+
"learning_rate": 0.00018367657050516163,
|
| 505 |
+
"loss": 0.463,
|
| 506 |
+
"step": 7100
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.847457627118644,
|
| 510 |
+
"grad_norm": 1.921875,
|
| 511 |
+
"learning_rate": 0.00018315093211213962,
|
| 512 |
+
"loss": 0.4597,
|
| 513 |
+
"step": 7200
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.8592278719397364,
|
| 517 |
+
"grad_norm": 6.875,
|
| 518 |
+
"learning_rate": 0.00018261774134023471,
|
| 519 |
+
"loss": 0.4571,
|
| 520 |
+
"step": 7300
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.8709981167608286,
|
| 524 |
+
"grad_norm": 11.1875,
|
| 525 |
+
"learning_rate": 0.00018207704661775216,
|
| 526 |
+
"loss": 0.4545,
|
| 527 |
+
"step": 7400
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.882768361581921,
|
| 531 |
+
"grad_norm": 2.140625,
|
| 532 |
+
"learning_rate": 0.00018152889705456117,
|
| 533 |
+
"loss": 0.4429,
|
| 534 |
+
"step": 7500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.8945386064030132,
|
| 538 |
+
"grad_norm": 4.84375,
|
| 539 |
+
"learning_rate": 0.00018097334243763453,
|
| 540 |
+
"loss": 0.4668,
|
| 541 |
+
"step": 7600
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.9063088512241054,
|
| 545 |
+
"grad_norm": 13.0,
|
| 546 |
+
"learning_rate": 0.00018041043322652626,
|
| 547 |
+
"loss": 0.4599,
|
| 548 |
+
"step": 7700
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.9180790960451978,
|
| 552 |
+
"grad_norm": 6.46875,
|
| 553 |
+
"learning_rate": 0.0001798402205487888,
|
| 554 |
+
"loss": 0.4561,
|
| 555 |
+
"step": 7800
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.92984934086629,
|
| 559 |
+
"grad_norm": 10.75,
|
| 560 |
+
"learning_rate": 0.00017926275619532906,
|
| 561 |
+
"loss": 0.4467,
|
| 562 |
+
"step": 7900
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.9416195856873822,
|
| 566 |
+
"grad_norm": 16.125,
|
| 567 |
+
"learning_rate": 0.00017867809261570448,
|
| 568 |
+
"loss": 0.4487,
|
| 569 |
+
"step": 8000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.9533898305084746,
|
| 573 |
+
"grad_norm": 8.4375,
|
| 574 |
+
"learning_rate": 0.00017808628291335912,
|
| 575 |
+
"loss": 0.4647,
|
| 576 |
+
"step": 8100
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.9651600753295668,
|
| 580 |
+
"grad_norm": 9.125,
|
| 581 |
+
"learning_rate": 0.0001774873808408004,
|
| 582 |
+
"loss": 0.4515,
|
| 583 |
+
"step": 8200
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.9769303201506592,
|
| 587 |
+
"grad_norm": 27.375,
|
| 588 |
+
"learning_rate": 0.00017688144079471707,
|
| 589 |
+
"loss": 0.4537,
|
| 590 |
+
"step": 8300
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.9887005649717514,
|
| 594 |
+
"grad_norm": 1.9765625,
|
| 595 |
+
"learning_rate": 0.0001762685178110382,
|
| 596 |
+
"loss": 0.45,
|
| 597 |
+
"step": 8400
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.0004708097928436,
|
| 601 |
+
"grad_norm": 29.75,
|
| 602 |
+
"learning_rate": 0.0001756486675599347,
|
| 603 |
+
"loss": 0.4565,
|
| 604 |
+
"step": 8500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.012241054613936,
|
| 608 |
+
"grad_norm": 1.84375,
|
| 609 |
+
"learning_rate": 0.00017502194634076278,
|
| 610 |
+
"loss": 0.4486,
|
| 611 |
+
"step": 8600
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.0240112994350283,
|
| 615 |
+
"grad_norm": 4.3125,
|
| 616 |
+
"learning_rate": 0.00017438841107695046,
|
| 617 |
+
"loss": 0.4556,
|
| 618 |
+
"step": 8700
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.0357815442561205,
|
| 622 |
+
"grad_norm": 15.5625,
|
| 623 |
+
"learning_rate": 0.00017374811931082743,
|
| 624 |
+
"loss": 0.4599,
|
| 625 |
+
"step": 8800
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.0475517890772128,
|
| 629 |
+
"grad_norm": 3.78125,
|
| 630 |
+
"learning_rate": 0.0001731011291983986,
|
| 631 |
+
"loss": 0.4479,
|
| 632 |
+
"step": 8900
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.0593220338983051,
|
| 636 |
+
"grad_norm": 18.875,
|
| 637 |
+
"learning_rate": 0.00017244749950406186,
|
| 638 |
+
"loss": 0.4488,
|
| 639 |
+
"step": 9000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.0710922787193973,
|
| 643 |
+
"grad_norm": 9.75,
|
| 644 |
+
"learning_rate": 0.00017178728959527076,
|
| 645 |
+
"loss": 0.4458,
|
| 646 |
+
"step": 9100
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.0828625235404896,
|
| 650 |
+
"grad_norm": 5.0625,
|
| 651 |
+
"learning_rate": 0.00017112055943714234,
|
| 652 |
+
"loss": 0.4352,
|
| 653 |
+
"step": 9200
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.094632768361582,
|
| 657 |
+
"grad_norm": 15.0625,
|
| 658 |
+
"learning_rate": 0.00017044736958701058,
|
| 659 |
+
"loss": 0.4359,
|
| 660 |
+
"step": 9300
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.1064030131826743,
|
| 664 |
+
"grad_norm": 4.28125,
|
| 665 |
+
"learning_rate": 0.00016976778118892619,
|
| 666 |
+
"loss": 0.4564,
|
| 667 |
+
"step": 9400
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.1181732580037664,
|
| 671 |
+
"grad_norm": 3.703125,
|
| 672 |
+
"learning_rate": 0.00016908185596810292,
|
| 673 |
+
"loss": 0.4508,
|
| 674 |
+
"step": 9500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.1299435028248588,
|
| 678 |
+
"grad_norm": 11.1875,
|
| 679 |
+
"learning_rate": 0.00016838965622531157,
|
| 680 |
+
"loss": 0.4489,
|
| 681 |
+
"step": 9600
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.1417137476459511,
|
| 685 |
+
"grad_norm": 8.0625,
|
| 686 |
+
"learning_rate": 0.00016769124483122098,
|
| 687 |
+
"loss": 0.4513,
|
| 688 |
+
"step": 9700
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.1534839924670433,
|
| 692 |
+
"grad_norm": 16.5,
|
| 693 |
+
"learning_rate": 0.00016698668522068788,
|
| 694 |
+
"loss": 0.4562,
|
| 695 |
+
"step": 9800
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.1652542372881356,
|
| 699 |
+
"grad_norm": 5.75,
|
| 700 |
+
"learning_rate": 0.00016627604138699515,
|
| 701 |
+
"loss": 0.4606,
|
| 702 |
+
"step": 9900
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.177024482109228,
|
| 706 |
+
"grad_norm": 7.34375,
|
| 707 |
+
"learning_rate": 0.00016555937787603967,
|
| 708 |
+
"loss": 0.4457,
|
| 709 |
+
"step": 10000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.18879472693032,
|
| 713 |
+
"grad_norm": 10.3125,
|
| 714 |
+
"learning_rate": 0.00016483675978046946,
|
| 715 |
+
"loss": 0.447,
|
| 716 |
+
"step": 10100
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.2005649717514124,
|
| 720 |
+
"grad_norm": 19.75,
|
| 721 |
+
"learning_rate": 0.00016410825273377192,
|
| 722 |
+
"loss": 0.452,
|
| 723 |
+
"step": 10200
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.2123352165725048,
|
| 727 |
+
"grad_norm": 6.65625,
|
| 728 |
+
"learning_rate": 0.0001633739229043121,
|
| 729 |
+
"loss": 0.4449,
|
| 730 |
+
"step": 10300
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.2241054613935969,
|
| 734 |
+
"grad_norm": 31.0,
|
| 735 |
+
"learning_rate": 0.00016263383698932306,
|
| 736 |
+
"loss": 0.4555,
|
| 737 |
+
"step": 10400
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.2358757062146892,
|
| 741 |
+
"grad_norm": 7.1875,
|
| 742 |
+
"learning_rate": 0.00016188806220884786,
|
| 743 |
+
"loss": 0.4664,
|
| 744 |
+
"step": 10500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.2476459510357816,
|
| 748 |
+
"grad_norm": 10.4375,
|
| 749 |
+
"learning_rate": 0.00016113666629963413,
|
| 750 |
+
"loss": 0.4559,
|
| 751 |
+
"step": 10600
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.2594161958568737,
|
| 755 |
+
"grad_norm": 8.4375,
|
| 756 |
+
"learning_rate": 0.00016037971750898172,
|
| 757 |
+
"loss": 0.4493,
|
| 758 |
+
"step": 10700
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.271186440677966,
|
| 762 |
+
"grad_norm": 8.875,
|
| 763 |
+
"learning_rate": 0.00015961728458854397,
|
| 764 |
+
"loss": 0.4556,
|
| 765 |
+
"step": 10800
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.2829566854990584,
|
| 769 |
+
"grad_norm": 3.296875,
|
| 770 |
+
"learning_rate": 0.00015884943678808324,
|
| 771 |
+
"loss": 0.4399,
|
| 772 |
+
"step": 10900
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.2947269303201507,
|
| 776 |
+
"grad_norm": 6.9375,
|
| 777 |
+
"learning_rate": 0.00015807624384918116,
|
| 778 |
+
"loss": 0.4572,
|
| 779 |
+
"step": 11000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.3064971751412429,
|
| 783 |
+
"grad_norm": 4.5625,
|
| 784 |
+
"learning_rate": 0.00015729777599890395,
|
| 785 |
+
"loss": 0.4358,
|
| 786 |
+
"step": 11100
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.3182674199623352,
|
| 790 |
+
"grad_norm": 10.625,
|
| 791 |
+
"learning_rate": 0.00015651410394342422,
|
| 792 |
+
"loss": 0.4469,
|
| 793 |
+
"step": 11200
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.3300376647834276,
|
| 797 |
+
"grad_norm": 8.9375,
|
| 798 |
+
"learning_rate": 0.00015572529886159872,
|
| 799 |
+
"loss": 0.4488,
|
| 800 |
+
"step": 11300
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.34180790960452,
|
| 804 |
+
"grad_norm": 9.0,
|
| 805 |
+
"learning_rate": 0.00015493143239850329,
|
| 806 |
+
"loss": 0.4443,
|
| 807 |
+
"step": 11400
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.353578154425612,
|
| 811 |
+
"grad_norm": 7.1875,
|
| 812 |
+
"learning_rate": 0.00015413257665892566,
|
| 813 |
+
"loss": 0.4427,
|
| 814 |
+
"step": 11500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.3653483992467044,
|
| 818 |
+
"grad_norm": 16.5,
|
| 819 |
+
"learning_rate": 0.00015332880420081647,
|
| 820 |
+
"loss": 0.468,
|
| 821 |
+
"step": 11600
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.3771186440677967,
|
| 825 |
+
"grad_norm": 2.390625,
|
| 826 |
+
"learning_rate": 0.00015252018802869866,
|
| 827 |
+
"loss": 0.4567,
|
| 828 |
+
"step": 11700
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.3888888888888888,
|
| 832 |
+
"grad_norm": 3.421875,
|
| 833 |
+
"learning_rate": 0.000151706801587037,
|
| 834 |
+
"loss": 0.4427,
|
| 835 |
+
"step": 11800
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.4006591337099812,
|
| 839 |
+
"grad_norm": 20.375,
|
| 840 |
+
"learning_rate": 0.00015088871875356723,
|
| 841 |
+
"loss": 0.454,
|
| 842 |
+
"step": 11900
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.4124293785310735,
|
| 846 |
+
"grad_norm": 12.25,
|
| 847 |
+
"learning_rate": 0.00015006601383258584,
|
| 848 |
+
"loss": 0.4537,
|
| 849 |
+
"step": 12000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.4241996233521657,
|
| 853 |
+
"grad_norm": 3.84375,
|
| 854 |
+
"learning_rate": 0.00014923876154820134,
|
| 855 |
+
"loss": 0.4528,
|
| 856 |
+
"step": 12100
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.435969868173258,
|
| 860 |
+
"grad_norm": 2.953125,
|
| 861 |
+
"learning_rate": 0.00014840703703754716,
|
| 862 |
+
"loss": 0.4481,
|
| 863 |
+
"step": 12200
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.4477401129943503,
|
| 867 |
+
"grad_norm": 1.9296875,
|
| 868 |
+
"learning_rate": 0.00014757091584395726,
|
| 869 |
+
"loss": 0.4497,
|
| 870 |
+
"step": 12300
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.4595103578154425,
|
| 874 |
+
"grad_norm": 3.375,
|
| 875 |
+
"learning_rate": 0.00014673047391010463,
|
| 876 |
+
"loss": 0.453,
|
| 877 |
+
"step": 12400
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.4712806026365348,
|
| 881 |
+
"grad_norm": 5.4375,
|
| 882 |
+
"learning_rate": 0.00014588578757110358,
|
| 883 |
+
"loss": 0.4604,
|
| 884 |
+
"step": 12500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.4830508474576272,
|
| 888 |
+
"grad_norm": 10.1875,
|
| 889 |
+
"learning_rate": 0.00014503693354757667,
|
| 890 |
+
"loss": 0.462,
|
| 891 |
+
"step": 12600
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.4948210922787193,
|
| 895 |
+
"grad_norm": 8.25,
|
| 896 |
+
"learning_rate": 0.000144183988938686,
|
| 897 |
+
"loss": 0.4348,
|
| 898 |
+
"step": 12700
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.5065913370998116,
|
| 902 |
+
"grad_norm": 3.6875,
|
| 903 |
+
"learning_rate": 0.00014332703121513088,
|
| 904 |
+
"loss": 0.4358,
|
| 905 |
+
"step": 12800
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.518361581920904,
|
| 909 |
+
"grad_norm": 7.75,
|
| 910 |
+
"learning_rate": 0.00014246613821211108,
|
| 911 |
+
"loss": 0.4546,
|
| 912 |
+
"step": 12900
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.530131826741996,
|
| 916 |
+
"grad_norm": 7.5625,
|
| 917 |
+
"learning_rate": 0.00014160138812225738,
|
| 918 |
+
"loss": 0.448,
|
| 919 |
+
"step": 13000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.5419020715630887,
|
| 923 |
+
"grad_norm": 1.875,
|
| 924 |
+
"learning_rate": 0.00014073285948852963,
|
| 925 |
+
"loss": 0.4559,
|
| 926 |
+
"step": 13100
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.5536723163841808,
|
| 930 |
+
"grad_norm": 4.46875,
|
| 931 |
+
"learning_rate": 0.00013986063119708275,
|
| 932 |
+
"loss": 0.4503,
|
| 933 |
+
"step": 13200
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.565442561205273,
|
| 937 |
+
"grad_norm": 1.796875,
|
| 938 |
+
"learning_rate": 0.00013898478247010177,
|
| 939 |
+
"loss": 0.4443,
|
| 940 |
+
"step": 13300
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.5772128060263655,
|
| 944 |
+
"grad_norm": 15.3125,
|
| 945 |
+
"learning_rate": 0.00013810539285860624,
|
| 946 |
+
"loss": 0.4559,
|
| 947 |
+
"step": 13400
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.5889830508474576,
|
| 951 |
+
"grad_norm": 3.390625,
|
| 952 |
+
"learning_rate": 0.000137222542235225,
|
| 953 |
+
"loss": 0.4448,
|
| 954 |
+
"step": 13500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.60075329566855,
|
| 958 |
+
"grad_norm": 8.6875,
|
| 959 |
+
"learning_rate": 0.00013633631078694127,
|
| 960 |
+
"loss": 0.4529,
|
| 961 |
+
"step": 13600
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.6125235404896423,
|
| 965 |
+
"grad_norm": 2.6875,
|
| 966 |
+
"learning_rate": 0.00013544677900780976,
|
| 967 |
+
"loss": 0.4487,
|
| 968 |
+
"step": 13700
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.6242937853107344,
|
| 972 |
+
"grad_norm": 6.40625,
|
| 973 |
+
"learning_rate": 0.0001345540276916455,
|
| 974 |
+
"loss": 0.4717,
|
| 975 |
+
"step": 13800
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.6360640301318268,
|
| 979 |
+
"grad_norm": 5.9375,
|
| 980 |
+
"learning_rate": 0.00013365813792468539,
|
| 981 |
+
"loss": 0.4464,
|
| 982 |
+
"step": 13900
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.6478342749529191,
|
| 986 |
+
"grad_norm": 20.25,
|
| 987 |
+
"learning_rate": 0.00013275919107822364,
|
| 988 |
+
"loss": 0.4522,
|
| 989 |
+
"step": 14000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.6596045197740112,
|
| 993 |
+
"grad_norm": 4.0,
|
| 994 |
+
"learning_rate": 0.0001318572688012209,
|
| 995 |
+
"loss": 0.4284,
|
| 996 |
+
"step": 14100
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.6713747645951036,
|
| 1000 |
+
"grad_norm": 15.5625,
|
| 1001 |
+
"learning_rate": 0.0001309524530128882,
|
| 1002 |
+
"loss": 0.436,
|
| 1003 |
+
"step": 14200
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.683145009416196,
|
| 1007 |
+
"grad_norm": 7.1875,
|
| 1008 |
+
"learning_rate": 0.0001300448258952466,
|
| 1009 |
+
"loss": 0.4545,
|
| 1010 |
+
"step": 14300
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.694915254237288,
|
| 1014 |
+
"grad_norm": 7.90625,
|
| 1015 |
+
"learning_rate": 0.00012913446988566273,
|
| 1016 |
+
"loss": 0.4605,
|
| 1017 |
+
"step": 14400
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.7066854990583804,
|
| 1021 |
+
"grad_norm": 8.3125,
|
| 1022 |
+
"learning_rate": 0.00012822146766936125,
|
| 1023 |
+
"loss": 0.4539,
|
| 1024 |
+
"step": 14500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.7184557438794728,
|
| 1028 |
+
"grad_norm": 4.75,
|
| 1029 |
+
"learning_rate": 0.00012730590217191479,
|
| 1030 |
+
"loss": 0.4354,
|
| 1031 |
+
"step": 14600
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.7302259887005649,
|
| 1035 |
+
"grad_norm": 9.1875,
|
| 1036 |
+
"learning_rate": 0.00012638785655171196,
|
| 1037 |
+
"loss": 0.4569,
|
| 1038 |
+
"step": 14700
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.7419962335216572,
|
| 1042 |
+
"grad_norm": 2.0,
|
| 1043 |
+
"learning_rate": 0.00012546741419240445,
|
| 1044 |
+
"loss": 0.438,
|
| 1045 |
+
"step": 14800
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.7537664783427496,
|
| 1049 |
+
"grad_norm": 4.5625,
|
| 1050 |
+
"learning_rate": 0.00012454465869533328,
|
| 1051 |
+
"loss": 0.4541,
|
| 1052 |
+
"step": 14900
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.7655367231638417,
|
| 1056 |
+
"grad_norm": 3.3125,
|
| 1057 |
+
"learning_rate": 0.0001236196738719357,
|
| 1058 |
+
"loss": 0.4317,
|
| 1059 |
+
"step": 15000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.777306967984934,
|
| 1063 |
+
"grad_norm": 16.625,
|
| 1064 |
+
"learning_rate": 0.00012269254373613287,
|
| 1065 |
+
"loss": 0.4593,
|
| 1066 |
+
"step": 15100
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.7890772128060264,
|
| 1070 |
+
"grad_norm": 7.1875,
|
| 1071 |
+
"learning_rate": 0.00012176335249669888,
|
| 1072 |
+
"loss": 0.431,
|
| 1073 |
+
"step": 15200
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.8008474576271185,
|
| 1077 |
+
"grad_norm": 11.125,
|
| 1078 |
+
"learning_rate": 0.00012083218454961237,
|
| 1079 |
+
"loss": 0.4362,
|
| 1080 |
+
"step": 15300
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.812617702448211,
|
| 1084 |
+
"grad_norm": 6.78125,
|
| 1085 |
+
"learning_rate": 0.00011989912447039113,
|
| 1086 |
+
"loss": 0.4351,
|
| 1087 |
+
"step": 15400
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.8243879472693032,
|
| 1091 |
+
"grad_norm": 14.625,
|
| 1092 |
+
"learning_rate": 0.00011896425700641042,
|
| 1093 |
+
"loss": 0.4485,
|
| 1094 |
+
"step": 15500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.8361581920903953,
|
| 1098 |
+
"grad_norm": 6.59375,
|
| 1099 |
+
"learning_rate": 0.00011802766706920533,
|
| 1100 |
+
"loss": 0.4403,
|
| 1101 |
+
"step": 15600
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.847928436911488,
|
| 1105 |
+
"grad_norm": 4.15625,
|
| 1106 |
+
"learning_rate": 0.00011708943972675866,
|
| 1107 |
+
"loss": 0.4438,
|
| 1108 |
+
"step": 15700
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.85969868173258,
|
| 1112 |
+
"grad_norm": 18.5,
|
| 1113 |
+
"learning_rate": 0.00011614966019577442,
|
| 1114 |
+
"loss": 0.4447,
|
| 1115 |
+
"step": 15800
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.8714689265536724,
|
| 1119 |
+
"grad_norm": 6.6875,
|
| 1120 |
+
"learning_rate": 0.00011520841383393774,
|
| 1121 |
+
"loss": 0.4475,
|
| 1122 |
+
"step": 15900
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.8832391713747647,
|
| 1126 |
+
"grad_norm": 7.6875,
|
| 1127 |
+
"learning_rate": 0.00011426578613216217,
|
| 1128 |
+
"loss": 0.4433,
|
| 1129 |
+
"step": 16000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.8950094161958568,
|
| 1133 |
+
"grad_norm": 6.625,
|
| 1134 |
+
"learning_rate": 0.00011332186270682473,
|
| 1135 |
+
"loss": 0.4472,
|
| 1136 |
+
"step": 16100
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.9067796610169492,
|
| 1140 |
+
"grad_norm": 5.46875,
|
| 1141 |
+
"learning_rate": 0.00011237672929198944,
|
| 1142 |
+
"loss": 0.4343,
|
| 1143 |
+
"step": 16200
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.9185499058380415,
|
| 1147 |
+
"grad_norm": 2.625,
|
| 1148 |
+
"learning_rate": 0.00011143047173162057,
|
| 1149 |
+
"loss": 0.4347,
|
| 1150 |
+
"step": 16300
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.9303201506591336,
|
| 1154 |
+
"grad_norm": 2.171875,
|
| 1155 |
+
"learning_rate": 0.00011048317597178561,
|
| 1156 |
+
"loss": 0.4326,
|
| 1157 |
+
"step": 16400
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.942090395480226,
|
| 1161 |
+
"grad_norm": 6.28125,
|
| 1162 |
+
"learning_rate": 0.00010953492805284882,
|
| 1163 |
+
"loss": 0.4489,
|
| 1164 |
+
"step": 16500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.9538606403013183,
|
| 1168 |
+
"grad_norm": 14.4375,
|
| 1169 |
+
"learning_rate": 0.0001085858141016566,
|
| 1170 |
+
"loss": 0.4524,
|
| 1171 |
+
"step": 16600
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.9656308851224105,
|
| 1175 |
+
"grad_norm": 9.75,
|
| 1176 |
+
"learning_rate": 0.00010763592032371482,
|
| 1177 |
+
"loss": 0.4456,
|
| 1178 |
+
"step": 16700
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.9774011299435028,
|
| 1182 |
+
"grad_norm": 2.734375,
|
| 1183 |
+
"learning_rate": 0.00010668533299535885,
|
| 1184 |
+
"loss": 0.4479,
|
| 1185 |
+
"step": 16800
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.9891713747645952,
|
| 1189 |
+
"grad_norm": 4.0625,
|
| 1190 |
+
"learning_rate": 0.00010573413845591748,
|
| 1191 |
+
"loss": 0.4495,
|
| 1192 |
+
"step": 16900
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 2.0009416195856873,
|
| 1196 |
+
"grad_norm": 11.0,
|
| 1197 |
+
"learning_rate": 0.00010478242309987088,
|
| 1198 |
+
"loss": 0.4326,
|
| 1199 |
+
"step": 17000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 2.01271186440678,
|
| 1203 |
+
"grad_norm": 7.28125,
|
| 1204 |
+
"learning_rate": 0.00010383027336900355,
|
| 1205 |
+
"loss": 0.4483,
|
| 1206 |
+
"step": 17100
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 2.024482109227872,
|
| 1210 |
+
"grad_norm": 10.3125,
|
| 1211 |
+
"learning_rate": 0.00010287777574455315,
|
| 1212 |
+
"loss": 0.4369,
|
| 1213 |
+
"step": 17200
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 2.036252354048964,
|
| 1217 |
+
"grad_norm": 8.625,
|
| 1218 |
+
"learning_rate": 0.00010192501673935557,
|
| 1219 |
+
"loss": 0.4239,
|
| 1220 |
+
"step": 17300
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 2.0480225988700567,
|
| 1224 |
+
"grad_norm": 9.1875,
|
| 1225 |
+
"learning_rate": 0.00010097208288998727,
|
| 1226 |
+
"loss": 0.4482,
|
| 1227 |
+
"step": 17400
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 2.059792843691149,
|
| 1231 |
+
"grad_norm": 5.34375,
|
| 1232 |
+
"learning_rate": 0.00010001906074890527,
|
| 1233 |
+
"loss": 0.4355,
|
| 1234 |
+
"step": 17500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 2.071563088512241,
|
| 1238 |
+
"grad_norm": 17.75,
|
| 1239 |
+
"learning_rate": 9.906603687658595e-05,
|
| 1240 |
+
"loss": 0.4389,
|
| 1241 |
+
"step": 17600
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 2.0833333333333335,
|
| 1245 |
+
"grad_norm": 1.75,
|
| 1246 |
+
"learning_rate": 9.811309783366291e-05,
|
| 1247 |
+
"loss": 0.4386,
|
| 1248 |
+
"step": 17700
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 2.0951035781544256,
|
| 1252 |
+
"grad_norm": 11.9375,
|
| 1253 |
+
"learning_rate": 9.716033017306491e-05,
|
| 1254 |
+
"loss": 0.4553,
|
| 1255 |
+
"step": 17800
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 2.1068738229755177,
|
| 1259 |
+
"grad_norm": 7.15625,
|
| 1260 |
+
"learning_rate": 9.620782043215454e-05,
|
| 1261 |
+
"loss": 0.4537,
|
| 1262 |
+
"step": 17900
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.1186440677966103,
|
| 1266 |
+
"grad_norm": 5.375,
|
| 1267 |
+
"learning_rate": 9.52556551248682e-05,
|
| 1268 |
+
"loss": 0.4468,
|
| 1269 |
+
"step": 18000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 2.1304143126177024,
|
| 1273 |
+
"grad_norm": 5.125,
|
| 1274 |
+
"learning_rate": 9.430392073385824e-05,
|
| 1275 |
+
"loss": 0.4374,
|
| 1276 |
+
"step": 18100
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 2.1421845574387945,
|
| 1280 |
+
"grad_norm": 4.03125,
|
| 1281 |
+
"learning_rate": 9.33527037026382e-05,
|
| 1282 |
+
"loss": 0.4431,
|
| 1283 |
+
"step": 18200
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 2.153954802259887,
|
| 1287 |
+
"grad_norm": 8.875,
|
| 1288 |
+
"learning_rate": 9.240209042773105e-05,
|
| 1289 |
+
"loss": 0.4401,
|
| 1290 |
+
"step": 18300
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 2.1657250470809792,
|
| 1294 |
+
"grad_norm": 5.25,
|
| 1295 |
+
"learning_rate": 9.145216725082219e-05,
|
| 1296 |
+
"loss": 0.447,
|
| 1297 |
+
"step": 18400
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 2.1774952919020714,
|
| 1301 |
+
"grad_norm": 10.125,
|
| 1302 |
+
"learning_rate": 9.050302045091731e-05,
|
| 1303 |
+
"loss": 0.4367,
|
| 1304 |
+
"step": 18500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 2.189265536723164,
|
| 1308 |
+
"grad_norm": 4.75,
|
| 1309 |
+
"learning_rate": 8.955473623650578e-05,
|
| 1310 |
+
"loss": 0.4416,
|
| 1311 |
+
"step": 18600
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 2.201035781544256,
|
| 1315 |
+
"grad_norm": 7.0,
|
| 1316 |
+
"learning_rate": 8.860740073773057e-05,
|
| 1317 |
+
"loss": 0.4385,
|
| 1318 |
+
"step": 18700
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 2.2128060263653486,
|
| 1322 |
+
"grad_norm": 1.8828125,
|
| 1323 |
+
"learning_rate": 8.766109999856553e-05,
|
| 1324 |
+
"loss": 0.4508,
|
| 1325 |
+
"step": 18800
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 2.2245762711864407,
|
| 1329 |
+
"grad_norm": 2.65625,
|
| 1330 |
+
"learning_rate": 8.671591996899974e-05,
|
| 1331 |
+
"loss": 0.4432,
|
| 1332 |
+
"step": 18900
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 2.236346516007533,
|
| 1336 |
+
"grad_norm": 4.0625,
|
| 1337 |
+
"learning_rate": 8.57719464972314e-05,
|
| 1338 |
+
"loss": 0.4423,
|
| 1339 |
+
"step": 19000
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 2.2481167608286254,
|
| 1343 |
+
"grad_norm": 2.1875,
|
| 1344 |
+
"learning_rate": 8.482926532187022e-05,
|
| 1345 |
+
"loss": 0.4476,
|
| 1346 |
+
"step": 19100
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 2.2598870056497176,
|
| 1350 |
+
"grad_norm": 16.25,
|
| 1351 |
+
"learning_rate": 8.388796206415004e-05,
|
| 1352 |
+
"loss": 0.4417,
|
| 1353 |
+
"step": 19200
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 2.2716572504708097,
|
| 1357 |
+
"grad_norm": 4.5,
|
| 1358 |
+
"learning_rate": 8.294812222015206e-05,
|
| 1359 |
+
"loss": 0.4419,
|
| 1360 |
+
"step": 19300
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 2.2834274952919023,
|
| 1364 |
+
"grad_norm": 3.609375,
|
| 1365 |
+
"learning_rate": 8.200983115303953e-05,
|
| 1366 |
+
"loss": 0.4184,
|
| 1367 |
+
"step": 19400
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 2.2951977401129944,
|
| 1371 |
+
"grad_norm": 1.515625,
|
| 1372 |
+
"learning_rate": 8.107317408530441e-05,
|
| 1373 |
+
"loss": 0.4314,
|
| 1374 |
+
"step": 19500
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 2.3069679849340865,
|
| 1378 |
+
"grad_norm": 3.671875,
|
| 1379 |
+
"learning_rate": 8.013823609102673e-05,
|
| 1380 |
+
"loss": 0.4603,
|
| 1381 |
+
"step": 19600
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 2.318738229755179,
|
| 1385 |
+
"grad_norm": 11.3125,
|
| 1386 |
+
"learning_rate": 7.920510208814779e-05,
|
| 1387 |
+
"loss": 0.4465,
|
| 1388 |
+
"step": 19700
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 2.330508474576271,
|
| 1392 |
+
"grad_norm": 23.625,
|
| 1393 |
+
"learning_rate": 7.827385683075701e-05,
|
| 1394 |
+
"loss": 0.4501,
|
| 1395 |
+
"step": 19800
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 2.3422787193973633,
|
| 1399 |
+
"grad_norm": 9.375,
|
| 1400 |
+
"learning_rate": 7.734458490139404e-05,
|
| 1401 |
+
"loss": 0.4451,
|
| 1402 |
+
"step": 19900
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 2.354048964218456,
|
| 1406 |
+
"grad_norm": 9.1875,
|
| 1407 |
+
"learning_rate": 7.641737070336649e-05,
|
| 1408 |
+
"loss": 0.455,
|
| 1409 |
+
"step": 20000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 2.365819209039548,
|
| 1413 |
+
"grad_norm": 5.46875,
|
| 1414 |
+
"learning_rate": 7.549229845308362e-05,
|
| 1415 |
+
"loss": 0.4444,
|
| 1416 |
+
"step": 20100
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 2.37758945386064,
|
| 1420 |
+
"grad_norm": 5.09375,
|
| 1421 |
+
"learning_rate": 7.456945217240721e-05,
|
| 1422 |
+
"loss": 0.4364,
|
| 1423 |
+
"step": 20200
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 2.3893596986817327,
|
| 1427 |
+
"grad_norm": 4.90625,
|
| 1428 |
+
"learning_rate": 7.36489156810202e-05,
|
| 1429 |
+
"loss": 0.4502,
|
| 1430 |
+
"step": 20300
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 2.401129943502825,
|
| 1434 |
+
"grad_norm": 2.8125,
|
| 1435 |
+
"learning_rate": 7.273077258881342e-05,
|
| 1436 |
+
"loss": 0.4303,
|
| 1437 |
+
"step": 20400
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 2.412900188323917,
|
| 1441 |
+
"grad_norm": 7.34375,
|
| 1442 |
+
"learning_rate": 7.181510628829158e-05,
|
| 1443 |
+
"loss": 0.4454,
|
| 1444 |
+
"step": 20500
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 2.4246704331450095,
|
| 1448 |
+
"grad_norm": 4.625,
|
| 1449 |
+
"learning_rate": 7.090199994699911e-05,
|
| 1450 |
+
"loss": 0.452,
|
| 1451 |
+
"step": 20600
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 2.4364406779661016,
|
| 1455 |
+
"grad_norm": 2.609375,
|
| 1456 |
+
"learning_rate": 6.999153649996595e-05,
|
| 1457 |
+
"loss": 0.4465,
|
| 1458 |
+
"step": 20700
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 2.4482109227871938,
|
| 1462 |
+
"grad_norm": 4.53125,
|
| 1463 |
+
"learning_rate": 6.9083798642175e-05,
|
| 1464 |
+
"loss": 0.4435,
|
| 1465 |
+
"step": 20800
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 2.4599811676082863,
|
| 1469 |
+
"grad_norm": 8.9375,
|
| 1470 |
+
"learning_rate": 6.81788688210512e-05,
|
| 1471 |
+
"loss": 0.421,
|
| 1472 |
+
"step": 20900
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 2.4717514124293785,
|
| 1476 |
+
"grad_norm": 16.375,
|
| 1477 |
+
"learning_rate": 6.727682922897282e-05,
|
| 1478 |
+
"loss": 0.4429,
|
| 1479 |
+
"step": 21000
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 2.4835216572504706,
|
| 1483 |
+
"grad_norm": 11.875,
|
| 1484 |
+
"learning_rate": 6.637776179580641e-05,
|
| 1485 |
+
"loss": 0.4327,
|
| 1486 |
+
"step": 21100
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 2.495291902071563,
|
| 1490 |
+
"grad_norm": 5.40625,
|
| 1491 |
+
"learning_rate": 6.548174818146514e-05,
|
| 1492 |
+
"loss": 0.4435,
|
| 1493 |
+
"step": 21200
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 2.5070621468926553,
|
| 1497 |
+
"grad_norm": 11.0,
|
| 1498 |
+
"learning_rate": 6.458886976849183e-05,
|
| 1499 |
+
"loss": 0.4429,
|
| 1500 |
+
"step": 21300
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 2.5188323917137474,
|
| 1504 |
+
"grad_norm": 2.109375,
|
| 1505 |
+
"learning_rate": 6.369920765466746e-05,
|
| 1506 |
+
"loss": 0.4286,
|
| 1507 |
+
"step": 21400
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 2.53060263653484,
|
| 1511 |
+
"grad_norm": 9.0625,
|
| 1512 |
+
"learning_rate": 6.28128426456449e-05,
|
| 1513 |
+
"loss": 0.4445,
|
| 1514 |
+
"step": 21500
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 2.542372881355932,
|
| 1518 |
+
"grad_norm": 15.375,
|
| 1519 |
+
"learning_rate": 6.19298552476098e-05,
|
| 1520 |
+
"loss": 0.4619,
|
| 1521 |
+
"step": 21600
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 2.554143126177024,
|
| 1525 |
+
"grad_norm": 2.03125,
|
| 1526 |
+
"learning_rate": 6.105032565996836e-05,
|
| 1527 |
+
"loss": 0.4357,
|
| 1528 |
+
"step": 21700
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 2.565913370998117,
|
| 1532 |
+
"grad_norm": 23.75,
|
| 1533 |
+
"learning_rate": 6.017433376806296e-05,
|
| 1534 |
+
"loss": 0.4436,
|
| 1535 |
+
"step": 21800
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 2.577683615819209,
|
| 1539 |
+
"grad_norm": 2.1875,
|
| 1540 |
+
"learning_rate": 5.9301959135916496e-05,
|
| 1541 |
+
"loss": 0.4567,
|
| 1542 |
+
"step": 21900
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 2.5894538606403015,
|
| 1546 |
+
"grad_norm": 4.09375,
|
| 1547 |
+
"learning_rate": 5.84332809990058e-05,
|
| 1548 |
+
"loss": 0.4446,
|
| 1549 |
+
"step": 22000
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 2.6012241054613936,
|
| 1553 |
+
"grad_norm": 6.03125,
|
| 1554 |
+
"learning_rate": 5.756837825706474e-05,
|
| 1555 |
+
"loss": 0.4351,
|
| 1556 |
+
"step": 22100
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 2.6129943502824857,
|
| 1560 |
+
"grad_norm": 3.578125,
|
| 1561 |
+
"learning_rate": 5.670732946691808e-05,
|
| 1562 |
+
"loss": 0.4467,
|
| 1563 |
+
"step": 22200
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 2.6247645951035783,
|
| 1567 |
+
"grad_norm": 8.125,
|
| 1568 |
+
"learning_rate": 5.5850212835346375e-05,
|
| 1569 |
+
"loss": 0.4417,
|
| 1570 |
+
"step": 22300
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 2.6365348399246704,
|
| 1574 |
+
"grad_norm": 11.9375,
|
| 1575 |
+
"learning_rate": 5.499710621198262e-05,
|
| 1576 |
+
"loss": 0.4498,
|
| 1577 |
+
"step": 22400
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 2.648305084745763,
|
| 1581 |
+
"grad_norm": 13.6875,
|
| 1582 |
+
"learning_rate": 5.414808708224135e-05,
|
| 1583 |
+
"loss": 0.4454,
|
| 1584 |
+
"step": 22500
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 2.660075329566855,
|
| 1588 |
+
"grad_norm": 2.359375,
|
| 1589 |
+
"learning_rate": 5.330323256028101e-05,
|
| 1590 |
+
"loss": 0.4596,
|
| 1591 |
+
"step": 22600
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 2.6718455743879472,
|
| 1595 |
+
"grad_norm": 6.6875,
|
| 1596 |
+
"learning_rate": 5.2462619381999635e-05,
|
| 1597 |
+
"loss": 0.4407,
|
| 1598 |
+
"step": 22700
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 2.68361581920904,
|
| 1602 |
+
"grad_norm": 3.046875,
|
| 1603 |
+
"learning_rate": 5.162632389806523e-05,
|
| 1604 |
+
"loss": 0.4448,
|
| 1605 |
+
"step": 22800
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 2.695386064030132,
|
| 1609 |
+
"grad_norm": 7.6875,
|
| 1610 |
+
"learning_rate": 5.079442206698121e-05,
|
| 1611 |
+
"loss": 0.4162,
|
| 1612 |
+
"step": 22900
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 2.707156308851224,
|
| 1616 |
+
"grad_norm": 12.1875,
|
| 1617 |
+
"learning_rate": 4.996698944818699e-05,
|
| 1618 |
+
"loss": 0.4467,
|
| 1619 |
+
"step": 23000
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 2.7189265536723166,
|
| 1623 |
+
"grad_norm": 6.03125,
|
| 1624 |
+
"learning_rate": 4.914410119519528e-05,
|
| 1625 |
+
"loss": 0.4397,
|
| 1626 |
+
"step": 23100
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 2.7306967984934087,
|
| 1630 |
+
"grad_norm": 2.1875,
|
| 1631 |
+
"learning_rate": 4.832583204876626e-05,
|
| 1632 |
+
"loss": 0.4434,
|
| 1633 |
+
"step": 23200
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 2.742467043314501,
|
| 1637 |
+
"grad_norm": 14.5,
|
| 1638 |
+
"learning_rate": 4.7512256330118645e-05,
|
| 1639 |
+
"loss": 0.4518,
|
| 1640 |
+
"step": 23300
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 2.7542372881355934,
|
| 1644 |
+
"grad_norm": 3.015625,
|
| 1645 |
+
"learning_rate": 4.670344793417967e-05,
|
| 1646 |
+
"loss": 0.4398,
|
| 1647 |
+
"step": 23400
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 2.7660075329566856,
|
| 1651 |
+
"grad_norm": 4.4375,
|
| 1652 |
+
"learning_rate": 4.589948032287334e-05,
|
| 1653 |
+
"loss": 0.4414,
|
| 1654 |
+
"step": 23500
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 2.7777777777777777,
|
| 1658 |
+
"grad_norm": 7.96875,
|
| 1659 |
+
"learning_rate": 4.5100426518447905e-05,
|
| 1660 |
+
"loss": 0.4478,
|
| 1661 |
+
"step": 23600
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 2.7895480225988702,
|
| 1665 |
+
"grad_norm": 5.5625,
|
| 1666 |
+
"learning_rate": 4.430635909684356e-05,
|
| 1667 |
+
"loss": 0.4585,
|
| 1668 |
+
"step": 23700
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 2.8013182674199624,
|
| 1672 |
+
"grad_norm": 6.03125,
|
| 1673 |
+
"learning_rate": 4.351735018110066e-05,
|
| 1674 |
+
"loss": 0.454,
|
| 1675 |
+
"step": 23800
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 2.8130885122410545,
|
| 1679 |
+
"grad_norm": 8.6875,
|
| 1680 |
+
"learning_rate": 4.273347143480877e-05,
|
| 1681 |
+
"loss": 0.4386,
|
| 1682 |
+
"step": 23900
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 2.824858757062147,
|
| 1686 |
+
"grad_norm": 2.390625,
|
| 1687 |
+
"learning_rate": 4.1954794055597756e-05,
|
| 1688 |
+
"loss": 0.4277,
|
| 1689 |
+
"step": 24000
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 2.836629001883239,
|
| 1693 |
+
"grad_norm": 10.1875,
|
| 1694 |
+
"learning_rate": 4.118138876867119e-05,
|
| 1695 |
+
"loss": 0.4457,
|
| 1696 |
+
"step": 24100
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 2.8483992467043313,
|
| 1700 |
+
"grad_norm": 2.84375,
|
| 1701 |
+
"learning_rate": 4.0413325820382373e-05,
|
| 1702 |
+
"loss": 0.4538,
|
| 1703 |
+
"step": 24200
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 2.860169491525424,
|
| 1707 |
+
"grad_norm": 1.7109375,
|
| 1708 |
+
"learning_rate": 3.965067497185416e-05,
|
| 1709 |
+
"loss": 0.4337,
|
| 1710 |
+
"step": 24300
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 2.871939736346516,
|
| 1714 |
+
"grad_norm": 18.375,
|
| 1715 |
+
"learning_rate": 3.889350549264275e-05,
|
| 1716 |
+
"loss": 0.4471,
|
| 1717 |
+
"step": 24400
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 2.883709981167608,
|
| 1721 |
+
"grad_norm": 9.125,
|
| 1722 |
+
"learning_rate": 3.814188615444602e-05,
|
| 1723 |
+
"loss": 0.4391,
|
| 1724 |
+
"step": 24500
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 2.8954802259887007,
|
| 1728 |
+
"grad_norm": 2.40625,
|
| 1729 |
+
"learning_rate": 3.739588522485736e-05,
|
| 1730 |
+
"loss": 0.4405,
|
| 1731 |
+
"step": 24600
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 2.907250470809793,
|
| 1735 |
+
"grad_norm": 3.5,
|
| 1736 |
+
"learning_rate": 3.665557046116491e-05,
|
| 1737 |
+
"loss": 0.4375,
|
| 1738 |
+
"step": 24700
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 2.919020715630885,
|
| 1742 |
+
"grad_norm": 3.703125,
|
| 1743 |
+
"learning_rate": 3.592100910419738e-05,
|
| 1744 |
+
"loss": 0.448,
|
| 1745 |
+
"step": 24800
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 2.9307909604519775,
|
| 1749 |
+
"grad_norm": 8.875,
|
| 1750 |
+
"learning_rate": 3.519226787221692e-05,
|
| 1751 |
+
"loss": 0.4356,
|
| 1752 |
+
"step": 24900
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 2.9425612052730696,
|
| 1756 |
+
"grad_norm": 12.125,
|
| 1757 |
+
"learning_rate": 3.446941295485905e-05,
|
| 1758 |
+
"loss": 0.4543,
|
| 1759 |
+
"step": 25000
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 2.9543314500941618,
|
| 1763 |
+
"grad_norm": 4.09375,
|
| 1764 |
+
"learning_rate": 3.375251000712094e-05,
|
| 1765 |
+
"loss": 0.4476,
|
| 1766 |
+
"step": 25100
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 2.9661016949152543,
|
| 1770 |
+
"grad_norm": 11.875,
|
| 1771 |
+
"learning_rate": 3.304162414339814e-05,
|
| 1772 |
+
"loss": 0.4476,
|
| 1773 |
+
"step": 25200
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 2.9778719397363465,
|
| 1777 |
+
"grad_norm": 10.625,
|
| 1778 |
+
"learning_rate": 3.23368199315704e-05,
|
| 1779 |
+
"loss": 0.4519,
|
| 1780 |
+
"step": 25300
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 2.9896421845574386,
|
| 1784 |
+
"grad_norm": 4.03125,
|
| 1785 |
+
"learning_rate": 3.163816138713708e-05,
|
| 1786 |
+
"loss": 0.4443,
|
| 1787 |
+
"step": 25400
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 3.001412429378531,
|
| 1791 |
+
"grad_norm": 15.0,
|
| 1792 |
+
"learning_rate": 3.094571196740299e-05,
|
| 1793 |
+
"loss": 0.4411,
|
| 1794 |
+
"step": 25500
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 3.0131826741996233,
|
| 1798 |
+
"grad_norm": 12.3125,
|
| 1799 |
+
"learning_rate": 3.0259534565714488e-05,
|
| 1800 |
+
"loss": 0.4403,
|
| 1801 |
+
"step": 25600
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 3.024952919020716,
|
| 1805 |
+
"grad_norm": 7.875,
|
| 1806 |
+
"learning_rate": 2.957969150574712e-05,
|
| 1807 |
+
"loss": 0.435,
|
| 1808 |
+
"step": 25700
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 3.036723163841808,
|
| 1812 |
+
"grad_norm": 12.5625,
|
| 1813 |
+
"learning_rate": 2.8906244535845072e-05,
|
| 1814 |
+
"loss": 0.444,
|
| 1815 |
+
"step": 25800
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 3.0484934086629,
|
| 1819 |
+
"grad_norm": 8.4375,
|
| 1820 |
+
"learning_rate": 2.823925482341252e-05,
|
| 1821 |
+
"loss": 0.4496,
|
| 1822 |
+
"step": 25900
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 3.0602636534839927,
|
| 1826 |
+
"grad_norm": 2.3125,
|
| 1827 |
+
"learning_rate": 2.7578782949358073e-05,
|
| 1828 |
+
"loss": 0.4441,
|
| 1829 |
+
"step": 26000
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 3.0720338983050848,
|
| 1833 |
+
"grad_norm": 10.125,
|
| 1834 |
+
"learning_rate": 2.692488890259235e-05,
|
| 1835 |
+
"loss": 0.4524,
|
| 1836 |
+
"step": 26100
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 3.083804143126177,
|
| 1840 |
+
"grad_norm": 1.96875,
|
| 1841 |
+
"learning_rate": 2.6277632074579338e-05,
|
| 1842 |
+
"loss": 0.4281,
|
| 1843 |
+
"step": 26200
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 3.0955743879472695,
|
| 1847 |
+
"grad_norm": 3.6875,
|
| 1848 |
+
"learning_rate": 2.5637071253942002e-05,
|
| 1849 |
+
"loss": 0.4364,
|
| 1850 |
+
"step": 26300
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 3.1073446327683616,
|
| 1854 |
+
"grad_norm": 2.203125,
|
| 1855 |
+
"learning_rate": 2.5003264621122802e-05,
|
| 1856 |
+
"loss": 0.4425,
|
| 1857 |
+
"step": 26400
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 3.1191148775894537,
|
| 1861 |
+
"grad_norm": 6.09375,
|
| 1862 |
+
"learning_rate": 2.4376269743099102e-05,
|
| 1863 |
+
"loss": 0.4227,
|
| 1864 |
+
"step": 26500
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 3.1308851224105463,
|
| 1868 |
+
"grad_norm": 6.65625,
|
| 1869 |
+
"learning_rate": 2.375614356815461e-05,
|
| 1870 |
+
"loss": 0.4615,
|
| 1871 |
+
"step": 26600
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 3.1426553672316384,
|
| 1875 |
+
"grad_norm": 5.375,
|
| 1876 |
+
"learning_rate": 2.314294242070706e-05,
|
| 1877 |
+
"loss": 0.4553,
|
| 1878 |
+
"step": 26700
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 3.1544256120527305,
|
| 1882 |
+
"grad_norm": 5.5,
|
| 1883 |
+
"learning_rate": 2.2536721996192168e-05,
|
| 1884 |
+
"loss": 0.4414,
|
| 1885 |
+
"step": 26800
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 3.166195856873823,
|
| 1889 |
+
"grad_norm": 2.4375,
|
| 1890 |
+
"learning_rate": 2.19375373560051e-05,
|
| 1891 |
+
"loss": 0.4525,
|
| 1892 |
+
"step": 26900
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 3.1779661016949152,
|
| 1896 |
+
"grad_norm": 14.625,
|
| 1897 |
+
"learning_rate": 2.1345442922499394e-05,
|
| 1898 |
+
"loss": 0.454,
|
| 1899 |
+
"step": 27000
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 3.1897363465160073,
|
| 1903 |
+
"grad_norm": 5.8125,
|
| 1904 |
+
"learning_rate": 2.076049247404388e-05,
|
| 1905 |
+
"loss": 0.4483,
|
| 1906 |
+
"step": 27100
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 3.2015065913371,
|
| 1910 |
+
"grad_norm": 9.1875,
|
| 1911 |
+
"learning_rate": 2.018273914013814e-05,
|
| 1912 |
+
"loss": 0.4437,
|
| 1913 |
+
"step": 27200
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 3.213276836158192,
|
| 1917 |
+
"grad_norm": 5.3125,
|
| 1918 |
+
"learning_rate": 1.9612235396586976e-05,
|
| 1919 |
+
"loss": 0.4423,
|
| 1920 |
+
"step": 27300
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 3.225047080979284,
|
| 1924 |
+
"grad_norm": 13.8125,
|
| 1925 |
+
"learning_rate": 1.904903306073401e-05,
|
| 1926 |
+
"loss": 0.4485,
|
| 1927 |
+
"step": 27400
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 3.2368173258003767,
|
| 1931 |
+
"grad_norm": 9.0625,
|
| 1932 |
+
"learning_rate": 1.8493183286755423e-05,
|
| 1933 |
+
"loss": 0.4435,
|
| 1934 |
+
"step": 27500
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 3.248587570621469,
|
| 1938 |
+
"grad_norm": 4.15625,
|
| 1939 |
+
"learning_rate": 1.794473656101364e-05,
|
| 1940 |
+
"loss": 0.4339,
|
| 1941 |
+
"step": 27600
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 3.2603578154425614,
|
| 1945 |
+
"grad_norm": 10.25,
|
| 1946 |
+
"learning_rate": 1.7403742697471824e-05,
|
| 1947 |
+
"loss": 0.4312,
|
| 1948 |
+
"step": 27700
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 3.2721280602636535,
|
| 1952 |
+
"grad_norm": 5.34375,
|
| 1953 |
+
"learning_rate": 1.687025083316942e-05,
|
| 1954 |
+
"loss": 0.4342,
|
| 1955 |
+
"step": 27800
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 3.2838983050847457,
|
| 1959 |
+
"grad_norm": 4.84375,
|
| 1960 |
+
"learning_rate": 1.6344309423759142e-05,
|
| 1961 |
+
"loss": 0.4368,
|
| 1962 |
+
"step": 27900
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 3.2956685499058382,
|
| 1966 |
+
"grad_norm": 13.625,
|
| 1967 |
+
"learning_rate": 1.5825966239105882e-05,
|
| 1968 |
+
"loss": 0.4312,
|
| 1969 |
+
"step": 28000
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 3.3074387947269304,
|
| 1973 |
+
"grad_norm": 10.25,
|
| 1974 |
+
"learning_rate": 1.5315268358947987e-05,
|
| 1975 |
+
"loss": 0.4367,
|
| 1976 |
+
"step": 28100
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 3.3192090395480225,
|
| 1980 |
+
"grad_norm": 3.546875,
|
| 1981 |
+
"learning_rate": 1.4812262168620939e-05,
|
| 1982 |
+
"loss": 0.4364,
|
| 1983 |
+
"step": 28200
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 3.330979284369115,
|
| 1987 |
+
"grad_norm": 8.1875,
|
| 1988 |
+
"learning_rate": 1.4316993354844444e-05,
|
| 1989 |
+
"loss": 0.4529,
|
| 1990 |
+
"step": 28300
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 3.342749529190207,
|
| 1994 |
+
"grad_norm": 3.0625,
|
| 1995 |
+
"learning_rate": 1.3829506901572853e-05,
|
| 1996 |
+
"loss": 0.4415,
|
| 1997 |
+
"step": 28400
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 3.3545197740112993,
|
| 2001 |
+
"grad_norm": 3.046875,
|
| 2002 |
+
"learning_rate": 1.3349847085909262e-05,
|
| 2003 |
+
"loss": 0.4292,
|
| 2004 |
+
"step": 28500
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 3.366290018832392,
|
| 2008 |
+
"grad_norm": 3.234375,
|
| 2009 |
+
"learning_rate": 1.2878057474084004e-05,
|
| 2010 |
+
"loss": 0.4493,
|
| 2011 |
+
"step": 28600
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 3.378060263653484,
|
| 2015 |
+
"grad_norm": 18.125,
|
| 2016 |
+
"learning_rate": 1.2414180917497743e-05,
|
| 2017 |
+
"loss": 0.4343,
|
| 2018 |
+
"step": 28700
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 3.389830508474576,
|
| 2022 |
+
"grad_norm": 2.078125,
|
| 2023 |
+
"learning_rate": 1.1958259548829153e-05,
|
| 2024 |
+
"loss": 0.4554,
|
| 2025 |
+
"step": 28800
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 3.4016007532956687,
|
| 2029 |
+
"grad_norm": 11.4375,
|
| 2030 |
+
"learning_rate": 1.1510334778208332e-05,
|
| 2031 |
+
"loss": 0.4496,
|
| 2032 |
+
"step": 28900
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 3.413370998116761,
|
| 2036 |
+
"grad_norm": 5.25,
|
| 2037 |
+
"learning_rate": 1.1070447289455632e-05,
|
| 2038 |
+
"loss": 0.4372,
|
| 2039 |
+
"step": 29000
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 3.425141242937853,
|
| 2043 |
+
"grad_norm": 14.9375,
|
| 2044 |
+
"learning_rate": 1.0638637036386279e-05,
|
| 2045 |
+
"loss": 0.4254,
|
| 2046 |
+
"step": 29100
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 3.4369114877589455,
|
| 2050 |
+
"grad_norm": 3.53125,
|
| 2051 |
+
"learning_rate": 1.0214943239181606e-05,
|
| 2052 |
+
"loss": 0.4426,
|
| 2053 |
+
"step": 29200
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 3.4486817325800376,
|
| 2057 |
+
"grad_norm": 6.0625,
|
| 2058 |
+
"learning_rate": 9.799404380826827e-06,
|
| 2059 |
+
"loss": 0.4429,
|
| 2060 |
+
"step": 29300
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 3.4604519774011298,
|
| 2064 |
+
"grad_norm": 6.0625,
|
| 2065 |
+
"learning_rate": 9.392058203615583e-06,
|
| 2066 |
+
"loss": 0.4349,
|
| 2067 |
+
"step": 29400
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 3.4722222222222223,
|
| 2071 |
+
"grad_norm": 2.921875,
|
| 2072 |
+
"learning_rate": 8.992941705722014e-06,
|
| 2073 |
+
"loss": 0.4466,
|
| 2074 |
+
"step": 29500
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 3.4839924670433144,
|
| 2078 |
+
"grad_norm": 7.03125,
|
| 2079 |
+
"learning_rate": 8.602091137840318e-06,
|
| 2080 |
+
"loss": 0.4381,
|
| 2081 |
+
"step": 29600
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 3.4957627118644066,
|
| 2085 |
+
"grad_norm": 3.46875,
|
| 2086 |
+
"learning_rate": 8.219541999892155e-06,
|
| 2087 |
+
"loss": 0.4479,
|
| 2088 |
+
"step": 29700
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 3.507532956685499,
|
| 2092 |
+
"grad_norm": 11.4375,
|
| 2093 |
+
"learning_rate": 7.845329037802229e-06,
|
| 2094 |
+
"loss": 0.4385,
|
| 2095 |
+
"step": 29800
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 3.5193032015065913,
|
| 2099 |
+
"grad_norm": 1.78125,
|
| 2100 |
+
"learning_rate": 7.479486240342604e-06,
|
| 2101 |
+
"loss": 0.4394,
|
| 2102 |
+
"step": 29900
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 3.5310734463276834,
|
| 2106 |
+
"grad_norm": 5.0625,
|
| 2107 |
+
"learning_rate": 7.122046836045348e-06,
|
| 2108 |
+
"loss": 0.4451,
|
| 2109 |
+
"step": 30000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 3.542843691148776,
|
| 2113 |
+
"grad_norm": 6.34375,
|
| 2114 |
+
"learning_rate": 6.773043290184644e-06,
|
| 2115 |
+
"loss": 0.4482,
|
| 2116 |
+
"step": 30100
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 3.554613935969868,
|
| 2120 |
+
"grad_norm": 1.1953125,
|
| 2121 |
+
"learning_rate": 6.432507301828028e-06,
|
| 2122 |
+
"loss": 0.4402,
|
| 2123 |
+
"step": 30200
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 3.56638418079096,
|
| 2127 |
+
"grad_norm": 5.65625,
|
| 2128 |
+
"learning_rate": 6.100469800957154e-06,
|
| 2129 |
+
"loss": 0.4393,
|
| 2130 |
+
"step": 30300
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 3.5781544256120528,
|
| 2134 |
+
"grad_norm": 6.5,
|
| 2135 |
+
"learning_rate": 5.7769609456585825e-06,
|
| 2136 |
+
"loss": 0.424,
|
| 2137 |
+
"step": 30400
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 3.589924670433145,
|
| 2141 |
+
"grad_norm": 2.15625,
|
| 2142 |
+
"learning_rate": 5.462010119384664e-06,
|
| 2143 |
+
"loss": 0.4531,
|
| 2144 |
+
"step": 30500
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 3.601694915254237,
|
| 2148 |
+
"grad_norm": 2.125,
|
| 2149 |
+
"learning_rate": 5.155645928284558e-06,
|
| 2150 |
+
"loss": 0.4494,
|
| 2151 |
+
"step": 30600
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 3.6134651600753296,
|
| 2155 |
+
"grad_norm": 6.96875,
|
| 2156 |
+
"learning_rate": 4.857896198606138e-06,
|
| 2157 |
+
"loss": 0.4412,
|
| 2158 |
+
"step": 30700
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 3.6252354048964217,
|
| 2162 |
+
"grad_norm": 15.8125,
|
| 2163 |
+
"learning_rate": 4.568787974168554e-06,
|
| 2164 |
+
"loss": 0.4351,
|
| 2165 |
+
"step": 30800
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 3.6370056497175143,
|
| 2169 |
+
"grad_norm": 4.46875,
|
| 2170 |
+
"learning_rate": 4.288347513905944e-06,
|
| 2171 |
+
"loss": 0.4331,
|
| 2172 |
+
"step": 30900
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 3.6487758945386064,
|
| 2176 |
+
"grad_norm": 7.3125,
|
| 2177 |
+
"learning_rate": 4.0166002894823864e-06,
|
| 2178 |
+
"loss": 0.4526,
|
| 2179 |
+
"step": 31000
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 3.6605461393596985,
|
| 2183 |
+
"grad_norm": 2.078125,
|
| 2184 |
+
"learning_rate": 3.7535709829783607e-06,
|
| 2185 |
+
"loss": 0.4527,
|
| 2186 |
+
"step": 31100
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 3.672316384180791,
|
| 2190 |
+
"grad_norm": 5.0,
|
| 2191 |
+
"learning_rate": 3.499283484648952e-06,
|
| 2192 |
+
"loss": 0.446,
|
| 2193 |
+
"step": 31200
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 3.684086629001883,
|
| 2197 |
+
"grad_norm": 9.0625,
|
| 2198 |
+
"learning_rate": 3.2537608907540206e-06,
|
| 2199 |
+
"loss": 0.4392,
|
| 2200 |
+
"step": 31300
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 3.695856873822976,
|
| 2204 |
+
"grad_norm": 9.1875,
|
| 2205 |
+
"learning_rate": 3.0170255014603244e-06,
|
| 2206 |
+
"loss": 0.4518,
|
| 2207 |
+
"step": 31400
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 3.707627118644068,
|
| 2211 |
+
"grad_norm": 6.5,
|
| 2212 |
+
"learning_rate": 2.789098818816105e-06,
|
| 2213 |
+
"loss": 0.4497,
|
| 2214 |
+
"step": 31500
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 3.71939736346516,
|
| 2218 |
+
"grad_norm": 8.25,
|
| 2219 |
+
"learning_rate": 2.5700015447981507e-06,
|
| 2220 |
+
"loss": 0.4575,
|
| 2221 |
+
"step": 31600
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 3.7311676082862526,
|
| 2225 |
+
"grad_norm": 5.40625,
|
| 2226 |
+
"learning_rate": 2.359753579431423e-06,
|
| 2227 |
+
"loss": 0.4382,
|
| 2228 |
+
"step": 31700
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 3.7429378531073447,
|
| 2232 |
+
"grad_norm": 6.0,
|
| 2233 |
+
"learning_rate": 2.158374018981635e-06,
|
| 2234 |
+
"loss": 0.4399,
|
| 2235 |
+
"step": 31800
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 3.754708097928437,
|
| 2239 |
+
"grad_norm": 3.984375,
|
| 2240 |
+
"learning_rate": 1.9658811542207835e-06,
|
| 2241 |
+
"loss": 0.4402,
|
| 2242 |
+
"step": 31900
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 3.7664783427495294,
|
| 2246 |
+
"grad_norm": 8.5,
|
| 2247 |
+
"learning_rate": 1.7822924687658227e-06,
|
| 2248 |
+
"loss": 0.4379,
|
| 2249 |
+
"step": 32000
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 3.7782485875706215,
|
| 2253 |
+
"grad_norm": 6.5,
|
| 2254 |
+
"learning_rate": 1.60762463749069e-06,
|
| 2255 |
+
"loss": 0.4482,
|
| 2256 |
+
"step": 32100
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 3.7900188323917137,
|
| 2260 |
+
"grad_norm": 3.546875,
|
| 2261 |
+
"learning_rate": 1.4418935250117837e-06,
|
| 2262 |
+
"loss": 0.4401,
|
| 2263 |
+
"step": 32200
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 3.8017890772128062,
|
| 2267 |
+
"grad_norm": 7.53125,
|
| 2268 |
+
"learning_rate": 1.2851141842469937e-06,
|
| 2269 |
+
"loss": 0.4488,
|
| 2270 |
+
"step": 32300
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 3.8135593220338984,
|
| 2274 |
+
"grad_norm": 2.796875,
|
| 2275 |
+
"learning_rate": 1.1373008550485175e-06,
|
| 2276 |
+
"loss": 0.4399,
|
| 2277 |
+
"step": 32400
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 3.8253295668549905,
|
| 2281 |
+
"grad_norm": 8.75,
|
| 2282 |
+
"learning_rate": 9.9846696290945e-07,
|
| 2283 |
+
"loss": 0.4341,
|
| 2284 |
+
"step": 32500
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 3.837099811676083,
|
| 2288 |
+
"grad_norm": 16.875,
|
| 2289 |
+
"learning_rate": 8.686251177444371e-07,
|
| 2290 |
+
"loss": 0.4489,
|
| 2291 |
+
"step": 32600
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 3.848870056497175,
|
| 2295 |
+
"grad_norm": 4.625,
|
| 2296 |
+
"learning_rate": 7.477871127442803e-07,
|
| 2297 |
+
"loss": 0.4342,
|
| 2298 |
+
"step": 32700
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 3.8606403013182673,
|
| 2302 |
+
"grad_norm": 6.8125,
|
| 2303 |
+
"learning_rate": 6.35963923304872e-07,
|
| 2304 |
+
"loss": 0.427,
|
| 2305 |
+
"step": 32800
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 3.87241054613936,
|
| 2309 |
+
"grad_norm": 4.125,
|
| 2310 |
+
"learning_rate": 5.331657060302586e-07,
|
| 2311 |
+
"loss": 0.4401,
|
| 2312 |
+
"step": 32900
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 3.884180790960452,
|
| 2316 |
+
"grad_norm": 18.0,
|
| 2317 |
+
"learning_rate": 4.3940179781019055e-07,
|
| 2318 |
+
"loss": 0.4432,
|
| 2319 |
+
"step": 33000
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 3.895951035781544,
|
| 2323 |
+
"grad_norm": 2.328125,
|
| 2324 |
+
"learning_rate": 3.546807149720555e-07,
|
| 2325 |
+
"loss": 0.439,
|
| 2326 |
+
"step": 33100
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 3.9077212806026367,
|
| 2330 |
+
"grad_norm": 3.171875,
|
| 2331 |
+
"learning_rate": 2.7901015250737515e-07,
|
| 2332 |
+
"loss": 0.4474,
|
| 2333 |
+
"step": 33200
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 3.919491525423729,
|
| 2337 |
+
"grad_norm": 7.40625,
|
| 2338 |
+
"learning_rate": 2.123969833728534e-07,
|
| 2339 |
+
"loss": 0.4568,
|
| 2340 |
+
"step": 33300
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 3.931261770244821,
|
| 2344 |
+
"grad_norm": 7.65625,
|
| 2345 |
+
"learning_rate": 1.5484725786619748e-07,
|
| 2346 |
+
"loss": 0.4505,
|
| 2347 |
+
"step": 33400
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 3.9430320150659135,
|
| 2351 |
+
"grad_norm": 11.125,
|
| 2352 |
+
"learning_rate": 1.0636620307650225e-07,
|
| 2353 |
+
"loss": 0.4382,
|
| 2354 |
+
"step": 33500
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 3.9548022598870056,
|
| 2358 |
+
"grad_norm": 3.703125,
|
| 2359 |
+
"learning_rate": 6.695822240952998e-08,
|
| 2360 |
+
"loss": 0.4416,
|
| 2361 |
+
"step": 33600
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 3.9665725047080977,
|
| 2365 |
+
"grad_norm": 1.734375,
|
| 2366 |
+
"learning_rate": 3.6626895187785726e-08,
|
| 2367 |
+
"loss": 0.4507,
|
| 2368 |
+
"step": 33700
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 3.9783427495291903,
|
| 2372 |
+
"grad_norm": 8.875,
|
| 2373 |
+
"learning_rate": 1.5374976325366332e-08,
|
| 2374 |
+
"loss": 0.4207,
|
| 2375 |
+
"step": 33800
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 3.9901129943502824,
|
| 2379 |
+
"grad_norm": 8.5,
|
| 2380 |
+
"learning_rate": 3.204396077749472e-09,
|
| 2381 |
+
"loss": 0.4692,
|
| 2382 |
+
"step": 33900
|
| 2383 |
+
}
|
| 2384 |
+
],
|
| 2385 |
+
"logging_steps": 100,
|
| 2386 |
+
"max_steps": 33984,
|
| 2387 |
+
"num_input_tokens_seen": 0,
|
| 2388 |
+
"num_train_epochs": 4,
|
| 2389 |
+
"save_steps": 4000,
|
| 2390 |
+
"stateful_callbacks": {
|
| 2391 |
+
"TrainerControl": {
|
| 2392 |
+
"args": {
|
| 2393 |
+
"should_epoch_stop": false,
|
| 2394 |
+
"should_evaluate": false,
|
| 2395 |
+
"should_log": false,
|
| 2396 |
+
"should_save": true,
|
| 2397 |
+
"should_training_stop": true
|
| 2398 |
+
},
|
| 2399 |
+
"attributes": {}
|
| 2400 |
+
}
|
| 2401 |
+
},
|
| 2402 |
+
"total_flos": 1.2131684208845605e+19,
|
| 2403 |
+
"train_batch_size": 16,
|
| 2404 |
+
"trial_name": null,
|
| 2405 |
+
"trial_params": null
|
| 2406 |
+
}
|