Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2702 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2702 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3800,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.02631578947368421,
|
| 13 |
+
"grad_norm": 9.638947486877441,
|
| 14 |
+
"learning_rate": 5.263157894736842e-06,
|
| 15 |
+
"loss": 1.4471,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.05263157894736842,
|
| 20 |
+
"grad_norm": 9.498916625976562,
|
| 21 |
+
"learning_rate": 1.0526315789473684e-05,
|
| 22 |
+
"loss": 1.2214,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.07894736842105263,
|
| 27 |
+
"grad_norm": 2.470106601715088,
|
| 28 |
+
"learning_rate": 1.5789473684210526e-05,
|
| 29 |
+
"loss": 0.6744,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.10526315789473684,
|
| 34 |
+
"grad_norm": 2.844433546066284,
|
| 35 |
+
"learning_rate": 2.105263157894737e-05,
|
| 36 |
+
"loss": 0.4362,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.13157894736842105,
|
| 41 |
+
"grad_norm": 1.874556303024292,
|
| 42 |
+
"learning_rate": 2.6315789473684212e-05,
|
| 43 |
+
"loss": 0.3136,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.15789473684210525,
|
| 48 |
+
"grad_norm": 1.6943750381469727,
|
| 49 |
+
"learning_rate": 3.157894736842105e-05,
|
| 50 |
+
"loss": 0.2589,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.18421052631578946,
|
| 55 |
+
"grad_norm": 2.1084232330322266,
|
| 56 |
+
"learning_rate": 3.6842105263157895e-05,
|
| 57 |
+
"loss": 0.2282,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.21052631578947367,
|
| 62 |
+
"grad_norm": 1.509325623512268,
|
| 63 |
+
"learning_rate": 4.210526315789474e-05,
|
| 64 |
+
"loss": 0.1952,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.23684210526315788,
|
| 69 |
+
"grad_norm": 2.0447444915771484,
|
| 70 |
+
"learning_rate": 4.736842105263158e-05,
|
| 71 |
+
"loss": 0.1842,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.2631578947368421,
|
| 76 |
+
"grad_norm": 1.149940013885498,
|
| 77 |
+
"learning_rate": 5.2631578947368424e-05,
|
| 78 |
+
"loss": 0.1915,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.2894736842105263,
|
| 83 |
+
"grad_norm": 1.1014671325683594,
|
| 84 |
+
"learning_rate": 5.789473684210527e-05,
|
| 85 |
+
"loss": 0.1688,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.3157894736842105,
|
| 90 |
+
"grad_norm": 1.9964191913604736,
|
| 91 |
+
"learning_rate": 6.31578947368421e-05,
|
| 92 |
+
"loss": 0.1494,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.34210526315789475,
|
| 97 |
+
"grad_norm": 0.7053777575492859,
|
| 98 |
+
"learning_rate": 6.842105263157895e-05,
|
| 99 |
+
"loss": 0.1418,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.3684210526315789,
|
| 104 |
+
"grad_norm": 1.077986240386963,
|
| 105 |
+
"learning_rate": 7.368421052631579e-05,
|
| 106 |
+
"loss": 0.1442,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.39473684210526316,
|
| 111 |
+
"grad_norm": 1.095324993133545,
|
| 112 |
+
"learning_rate": 7.894736842105263e-05,
|
| 113 |
+
"loss": 0.1332,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.42105263157894735,
|
| 118 |
+
"grad_norm": 0.9210699200630188,
|
| 119 |
+
"learning_rate": 8.421052631578948e-05,
|
| 120 |
+
"loss": 0.1322,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.4473684210526316,
|
| 125 |
+
"grad_norm": 1.3079745769500732,
|
| 126 |
+
"learning_rate": 8.947368421052632e-05,
|
| 127 |
+
"loss": 0.1204,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.47368421052631576,
|
| 132 |
+
"grad_norm": 1.196655035018921,
|
| 133 |
+
"learning_rate": 9.473684210526316e-05,
|
| 134 |
+
"loss": 0.1227,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.5,
|
| 139 |
+
"grad_norm": 0.8218312859535217,
|
| 140 |
+
"learning_rate": 0.0001,
|
| 141 |
+
"loss": 0.1215,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.5263157894736842,
|
| 146 |
+
"grad_norm": 0.9372950196266174,
|
| 147 |
+
"learning_rate": 9.999810668616086e-05,
|
| 148 |
+
"loss": 0.114,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.5526315789473685,
|
| 153 |
+
"grad_norm": 1.0510334968566895,
|
| 154 |
+
"learning_rate": 9.999242688802886e-05,
|
| 155 |
+
"loss": 0.1101,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.5789473684210527,
|
| 160 |
+
"grad_norm": 0.7145567536354065,
|
| 161 |
+
"learning_rate": 9.998296103574967e-05,
|
| 162 |
+
"loss": 0.0964,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.6052631578947368,
|
| 167 |
+
"grad_norm": 0.7895988821983337,
|
| 168 |
+
"learning_rate": 9.996970984619641e-05,
|
| 169 |
+
"loss": 0.0948,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.631578947368421,
|
| 174 |
+
"grad_norm": 0.8074854612350464,
|
| 175 |
+
"learning_rate": 9.995267432291555e-05,
|
| 176 |
+
"loss": 0.1025,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.6578947368421053,
|
| 181 |
+
"grad_norm": 0.9896809458732605,
|
| 182 |
+
"learning_rate": 9.993185575605073e-05,
|
| 183 |
+
"loss": 0.0954,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.6842105263157895,
|
| 188 |
+
"grad_norm": 1.2382564544677734,
|
| 189 |
+
"learning_rate": 9.990725572224521e-05,
|
| 190 |
+
"loss": 0.0965,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.7105263157894737,
|
| 195 |
+
"grad_norm": 0.6677968502044678,
|
| 196 |
+
"learning_rate": 9.987887608452235e-05,
|
| 197 |
+
"loss": 0.104,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.7368421052631579,
|
| 202 |
+
"grad_norm": 0.8802872896194458,
|
| 203 |
+
"learning_rate": 9.984671899214457e-05,
|
| 204 |
+
"loss": 0.0936,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.7631578947368421,
|
| 209 |
+
"grad_norm": 0.5323192477226257,
|
| 210 |
+
"learning_rate": 9.981078688045062e-05,
|
| 211 |
+
"loss": 0.0937,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.7894736842105263,
|
| 216 |
+
"grad_norm": 0.43582749366760254,
|
| 217 |
+
"learning_rate": 9.977108247067108e-05,
|
| 218 |
+
"loss": 0.0913,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.8157894736842105,
|
| 223 |
+
"grad_norm": 0.5718427896499634,
|
| 224 |
+
"learning_rate": 9.972760876972226e-05,
|
| 225 |
+
"loss": 0.0913,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.8421052631578947,
|
| 230 |
+
"grad_norm": 0.5905811190605164,
|
| 231 |
+
"learning_rate": 9.968036906997855e-05,
|
| 232 |
+
"loss": 0.091,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.868421052631579,
|
| 237 |
+
"grad_norm": 0.5551950931549072,
|
| 238 |
+
"learning_rate": 9.962936694902307e-05,
|
| 239 |
+
"loss": 0.0827,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.8947368421052632,
|
| 244 |
+
"grad_norm": 0.37185293436050415,
|
| 245 |
+
"learning_rate": 9.957460626937664e-05,
|
| 246 |
+
"loss": 0.0768,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.9210526315789473,
|
| 251 |
+
"grad_norm": 0.7125674486160278,
|
| 252 |
+
"learning_rate": 9.951609117820538e-05,
|
| 253 |
+
"loss": 0.0878,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.9473684210526315,
|
| 258 |
+
"grad_norm": 0.518326461315155,
|
| 259 |
+
"learning_rate": 9.945382610700657e-05,
|
| 260 |
+
"loss": 0.0841,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.9736842105263158,
|
| 265 |
+
"grad_norm": 0.40797701478004456,
|
| 266 |
+
"learning_rate": 9.938781577127306e-05,
|
| 267 |
+
"loss": 0.0899,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.0,
|
| 272 |
+
"grad_norm": 0.8398942351341248,
|
| 273 |
+
"learning_rate": 9.931806517013612e-05,
|
| 274 |
+
"loss": 0.082,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.0263157894736843,
|
| 279 |
+
"grad_norm": 0.5522546768188477,
|
| 280 |
+
"learning_rate": 9.92445795859869e-05,
|
| 281 |
+
"loss": 0.0833,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.0526315789473684,
|
| 286 |
+
"grad_norm": 0.7629371285438538,
|
| 287 |
+
"learning_rate": 9.916736458407632e-05,
|
| 288 |
+
"loss": 0.0782,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.0789473684210527,
|
| 293 |
+
"grad_norm": 0.6491877436637878,
|
| 294 |
+
"learning_rate": 9.908642601209366e-05,
|
| 295 |
+
"loss": 0.0795,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.1052631578947367,
|
| 300 |
+
"grad_norm": 0.5657724142074585,
|
| 301 |
+
"learning_rate": 9.900176999972366e-05,
|
| 302 |
+
"loss": 0.0763,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.131578947368421,
|
| 307 |
+
"grad_norm": 0.9348977208137512,
|
| 308 |
+
"learning_rate": 9.89134029581823e-05,
|
| 309 |
+
"loss": 0.0789,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.1578947368421053,
|
| 314 |
+
"grad_norm": 0.9060670733451843,
|
| 315 |
+
"learning_rate": 9.88213315797313e-05,
|
| 316 |
+
"loss": 0.0897,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.1842105263157894,
|
| 321 |
+
"grad_norm": 0.5024181008338928,
|
| 322 |
+
"learning_rate": 9.872556283717125e-05,
|
| 323 |
+
"loss": 0.0781,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.2105263157894737,
|
| 328 |
+
"grad_norm": 0.6639284491539001,
|
| 329 |
+
"learning_rate": 9.86261039833136e-05,
|
| 330 |
+
"loss": 0.0871,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.236842105263158,
|
| 335 |
+
"grad_norm": 0.572541356086731,
|
| 336 |
+
"learning_rate": 9.852296255043129e-05,
|
| 337 |
+
"loss": 0.0741,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.263157894736842,
|
| 342 |
+
"grad_norm": 0.8119587898254395,
|
| 343 |
+
"learning_rate": 9.841614634968843e-05,
|
| 344 |
+
"loss": 0.0732,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.2894736842105263,
|
| 349 |
+
"grad_norm": 0.45451977849006653,
|
| 350 |
+
"learning_rate": 9.830566347054868e-05,
|
| 351 |
+
"loss": 0.0734,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.3157894736842106,
|
| 356 |
+
"grad_norm": 0.48096764087677,
|
| 357 |
+
"learning_rate": 9.819152228016257e-05,
|
| 358 |
+
"loss": 0.0729,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.3421052631578947,
|
| 363 |
+
"grad_norm": 0.4565185606479645,
|
| 364 |
+
"learning_rate": 9.807373142273395e-05,
|
| 365 |
+
"loss": 0.0673,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.368421052631579,
|
| 370 |
+
"grad_norm": 0.7846236824989319,
|
| 371 |
+
"learning_rate": 9.795229981886521e-05,
|
| 372 |
+
"loss": 0.0687,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.3947368421052633,
|
| 377 |
+
"grad_norm": 0.46232283115386963,
|
| 378 |
+
"learning_rate": 9.782723666488181e-05,
|
| 379 |
+
"loss": 0.0718,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.4210526315789473,
|
| 384 |
+
"grad_norm": 0.5114668607711792,
|
| 385 |
+
"learning_rate": 9.769855143213575e-05,
|
| 386 |
+
"loss": 0.0739,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.4473684210526316,
|
| 391 |
+
"grad_norm": 0.7048826217651367,
|
| 392 |
+
"learning_rate": 9.756625386628832e-05,
|
| 393 |
+
"loss": 0.066,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.4736842105263157,
|
| 398 |
+
"grad_norm": 0.609493613243103,
|
| 399 |
+
"learning_rate": 9.743035398657201e-05,
|
| 400 |
+
"loss": 0.0744,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.5,
|
| 405 |
+
"grad_norm": 0.6096906065940857,
|
| 406 |
+
"learning_rate": 9.729086208503174e-05,
|
| 407 |
+
"loss": 0.0678,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.526315789473684,
|
| 412 |
+
"grad_norm": 0.8587449193000793,
|
| 413 |
+
"learning_rate": 9.714778872574541e-05,
|
| 414 |
+
"loss": 0.0656,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.5526315789473686,
|
| 419 |
+
"grad_norm": 0.576169490814209,
|
| 420 |
+
"learning_rate": 9.700114474402387e-05,
|
| 421 |
+
"loss": 0.0715,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.5789473684210527,
|
| 426 |
+
"grad_norm": 0.6562088131904602,
|
| 427 |
+
"learning_rate": 9.685094124559034e-05,
|
| 428 |
+
"loss": 0.0694,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.6052631578947367,
|
| 433 |
+
"grad_norm": 0.3795994818210602,
|
| 434 |
+
"learning_rate": 9.669718960573927e-05,
|
| 435 |
+
"loss": 0.0643,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.631578947368421,
|
| 440 |
+
"grad_norm": 0.5144829154014587,
|
| 441 |
+
"learning_rate": 9.653990146847499e-05,
|
| 442 |
+
"loss": 0.0713,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.6578947368421053,
|
| 447 |
+
"grad_norm": 0.34501057863235474,
|
| 448 |
+
"learning_rate": 9.637908874562978e-05,
|
| 449 |
+
"loss": 0.0655,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.6842105263157894,
|
| 454 |
+
"grad_norm": 0.3585371673107147,
|
| 455 |
+
"learning_rate": 9.621476361596177e-05,
|
| 456 |
+
"loss": 0.0634,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.7105263157894737,
|
| 461 |
+
"grad_norm": 0.4583841562271118,
|
| 462 |
+
"learning_rate": 9.604693852423268e-05,
|
| 463 |
+
"loss": 0.0603,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.736842105263158,
|
| 468 |
+
"grad_norm": 0.372306764125824,
|
| 469 |
+
"learning_rate": 9.58756261802652e-05,
|
| 470 |
+
"loss": 0.0606,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.763157894736842,
|
| 475 |
+
"grad_norm": 0.49868544936180115,
|
| 476 |
+
"learning_rate": 9.570083955798065e-05,
|
| 477 |
+
"loss": 0.0651,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.7894736842105263,
|
| 482 |
+
"grad_norm": 0.531895101070404,
|
| 483 |
+
"learning_rate": 9.552259189441626e-05,
|
| 484 |
+
"loss": 0.0708,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.8157894736842106,
|
| 489 |
+
"grad_norm": 0.5189070701599121,
|
| 490 |
+
"learning_rate": 9.534089668872274e-05,
|
| 491 |
+
"loss": 0.0688,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.8421052631578947,
|
| 496 |
+
"grad_norm": 0.36183297634124756,
|
| 497 |
+
"learning_rate": 9.515576770114199e-05,
|
| 498 |
+
"loss": 0.0684,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.868421052631579,
|
| 503 |
+
"grad_norm": 0.2646963894367218,
|
| 504 |
+
"learning_rate": 9.496721895196497e-05,
|
| 505 |
+
"loss": 0.0604,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.8947368421052633,
|
| 510 |
+
"grad_norm": 0.4395381510257721,
|
| 511 |
+
"learning_rate": 9.477526472046995e-05,
|
| 512 |
+
"loss": 0.0523,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.9210526315789473,
|
| 517 |
+
"grad_norm": 0.5579097867012024,
|
| 518 |
+
"learning_rate": 9.457991954384105e-05,
|
| 519 |
+
"loss": 0.0612,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.9473684210526314,
|
| 524 |
+
"grad_norm": 0.4455315172672272,
|
| 525 |
+
"learning_rate": 9.438119821606727e-05,
|
| 526 |
+
"loss": 0.0625,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.973684210526316,
|
| 531 |
+
"grad_norm": 0.5463271737098694,
|
| 532 |
+
"learning_rate": 9.417911578682229e-05,
|
| 533 |
+
"loss": 0.0605,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.0,
|
| 538 |
+
"grad_norm": 0.7862467765808105,
|
| 539 |
+
"learning_rate": 9.397368756032445e-05,
|
| 540 |
+
"loss": 0.0606,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.026315789473684,
|
| 545 |
+
"grad_norm": 0.48962733149528503,
|
| 546 |
+
"learning_rate": 9.376492909417795e-05,
|
| 547 |
+
"loss": 0.0565,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.0526315789473686,
|
| 552 |
+
"grad_norm": 0.37852731347084045,
|
| 553 |
+
"learning_rate": 9.35528561981945e-05,
|
| 554 |
+
"loss": 0.0622,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.0789473684210527,
|
| 559 |
+
"grad_norm": 0.37320488691329956,
|
| 560 |
+
"learning_rate": 9.333748493319603e-05,
|
| 561 |
+
"loss": 0.0602,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.1052631578947367,
|
| 566 |
+
"grad_norm": 0.5312328338623047,
|
| 567 |
+
"learning_rate": 9.311883160979844e-05,
|
| 568 |
+
"loss": 0.0522,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.1315789473684212,
|
| 573 |
+
"grad_norm": 0.5505270957946777,
|
| 574 |
+
"learning_rate": 9.289691278717623e-05,
|
| 575 |
+
"loss": 0.0544,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.1578947368421053,
|
| 580 |
+
"grad_norm": 0.6213037371635437,
|
| 581 |
+
"learning_rate": 9.267174527180853e-05,
|
| 582 |
+
"loss": 0.0647,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.1842105263157894,
|
| 587 |
+
"grad_norm": 0.5001884698867798,
|
| 588 |
+
"learning_rate": 9.244334611620629e-05,
|
| 589 |
+
"loss": 0.0573,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.2105263157894735,
|
| 594 |
+
"grad_norm": 0.41479626297950745,
|
| 595 |
+
"learning_rate": 9.221173261762073e-05,
|
| 596 |
+
"loss": 0.0576,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.236842105263158,
|
| 601 |
+
"grad_norm": 0.4636107385158539,
|
| 602 |
+
"learning_rate": 9.197692231673361e-05,
|
| 603 |
+
"loss": 0.0607,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.263157894736842,
|
| 608 |
+
"grad_norm": 0.4859972298145294,
|
| 609 |
+
"learning_rate": 9.173893299632856e-05,
|
| 610 |
+
"loss": 0.0561,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.2894736842105265,
|
| 615 |
+
"grad_norm": 0.3786041736602783,
|
| 616 |
+
"learning_rate": 9.149778267994457e-05,
|
| 617 |
+
"loss": 0.0594,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.3157894736842106,
|
| 622 |
+
"grad_norm": 0.5478907823562622,
|
| 623 |
+
"learning_rate": 9.12534896305109e-05,
|
| 624 |
+
"loss": 0.0547,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.3421052631578947,
|
| 629 |
+
"grad_norm": 0.3971192240715027,
|
| 630 |
+
"learning_rate": 9.100607234896397e-05,
|
| 631 |
+
"loss": 0.058,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.3684210526315788,
|
| 636 |
+
"grad_norm": 0.4420956075191498,
|
| 637 |
+
"learning_rate": 9.075554957284633e-05,
|
| 638 |
+
"loss": 0.0576,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.3947368421052633,
|
| 643 |
+
"grad_norm": 0.5636497139930725,
|
| 644 |
+
"learning_rate": 9.050194027488754e-05,
|
| 645 |
+
"loss": 0.0513,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.4210526315789473,
|
| 650 |
+
"grad_norm": 0.6197888255119324,
|
| 651 |
+
"learning_rate": 9.024526366156732e-05,
|
| 652 |
+
"loss": 0.0547,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.4473684210526314,
|
| 657 |
+
"grad_norm": 0.5714887976646423,
|
| 658 |
+
"learning_rate": 8.998553917166108e-05,
|
| 659 |
+
"loss": 0.0605,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.473684210526316,
|
| 664 |
+
"grad_norm": 0.3763100206851959,
|
| 665 |
+
"learning_rate": 8.972278647476764e-05,
|
| 666 |
+
"loss": 0.053,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.5,
|
| 671 |
+
"grad_norm": 0.4168098270893097,
|
| 672 |
+
"learning_rate": 8.945702546981969e-05,
|
| 673 |
+
"loss": 0.0587,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.526315789473684,
|
| 678 |
+
"grad_norm": 0.3482176959514618,
|
| 679 |
+
"learning_rate": 8.918827628357677e-05,
|
| 680 |
+
"loss": 0.053,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.5526315789473686,
|
| 685 |
+
"grad_norm": 0.4938332736492157,
|
| 686 |
+
"learning_rate": 8.891655926910103e-05,
|
| 687 |
+
"loss": 0.06,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.5789473684210527,
|
| 692 |
+
"grad_norm": 0.3451533317565918,
|
| 693 |
+
"learning_rate": 8.864189500421582e-05,
|
| 694 |
+
"loss": 0.0612,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.6052631578947367,
|
| 699 |
+
"grad_norm": 0.4287240505218506,
|
| 700 |
+
"learning_rate": 8.836430428994732e-05,
|
| 701 |
+
"loss": 0.054,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.6315789473684212,
|
| 706 |
+
"grad_norm": 0.6272952556610107,
|
| 707 |
+
"learning_rate": 8.808380814894912e-05,
|
| 708 |
+
"loss": 0.0555,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 2.6578947368421053,
|
| 713 |
+
"grad_norm": 0.46518048644065857,
|
| 714 |
+
"learning_rate": 8.780042782391028e-05,
|
| 715 |
+
"loss": 0.0608,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.6842105263157894,
|
| 720 |
+
"grad_norm": 0.7546030282974243,
|
| 721 |
+
"learning_rate": 8.751418477594645e-05,
|
| 722 |
+
"loss": 0.0566,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.7105263157894735,
|
| 727 |
+
"grad_norm": 0.6570419073104858,
|
| 728 |
+
"learning_rate": 8.722510068297454e-05,
|
| 729 |
+
"loss": 0.0565,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 2.736842105263158,
|
| 734 |
+
"grad_norm": 0.5381316542625427,
|
| 735 |
+
"learning_rate": 8.693319743807116e-05,
|
| 736 |
+
"loss": 0.0534,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 2.763157894736842,
|
| 741 |
+
"grad_norm": 0.566612958908081,
|
| 742 |
+
"learning_rate": 8.663849714781442e-05,
|
| 743 |
+
"loss": 0.05,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 2.7894736842105265,
|
| 748 |
+
"grad_norm": 0.49792104959487915,
|
| 749 |
+
"learning_rate": 8.634102213060984e-05,
|
| 750 |
+
"loss": 0.0501,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 2.8157894736842106,
|
| 755 |
+
"grad_norm": 0.39989131689071655,
|
| 756 |
+
"learning_rate": 8.60407949150001e-05,
|
| 757 |
+
"loss": 0.0541,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.8421052631578947,
|
| 762 |
+
"grad_norm": 0.3729485273361206,
|
| 763 |
+
"learning_rate": 8.573783823795889e-05,
|
| 764 |
+
"loss": 0.0568,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 2.8684210526315788,
|
| 769 |
+
"grad_norm": 0.3304407000541687,
|
| 770 |
+
"learning_rate": 8.543217504316896e-05,
|
| 771 |
+
"loss": 0.0508,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 2.8947368421052633,
|
| 776 |
+
"grad_norm": 0.3478791415691376,
|
| 777 |
+
"learning_rate": 8.512382847928461e-05,
|
| 778 |
+
"loss": 0.0544,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 2.9210526315789473,
|
| 783 |
+
"grad_norm": 0.35698530077934265,
|
| 784 |
+
"learning_rate": 8.48128218981785e-05,
|
| 785 |
+
"loss": 0.0522,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 2.9473684210526314,
|
| 790 |
+
"grad_norm": 0.37133410573005676,
|
| 791 |
+
"learning_rate": 8.44991788531732e-05,
|
| 792 |
+
"loss": 0.0488,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 2.973684210526316,
|
| 797 |
+
"grad_norm": 0.31634142994880676,
|
| 798 |
+
"learning_rate": 8.418292309725738e-05,
|
| 799 |
+
"loss": 0.0551,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.0,
|
| 804 |
+
"grad_norm": 0.5727193355560303,
|
| 805 |
+
"learning_rate": 8.386407858128706e-05,
|
| 806 |
+
"loss": 0.0497,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.026315789473684,
|
| 811 |
+
"grad_norm": 0.3365287482738495,
|
| 812 |
+
"learning_rate": 8.35426694521716e-05,
|
| 813 |
+
"loss": 0.048,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.0526315789473686,
|
| 818 |
+
"grad_norm": 0.26241013407707214,
|
| 819 |
+
"learning_rate": 8.321872005104509e-05,
|
| 820 |
+
"loss": 0.0517,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.0789473684210527,
|
| 825 |
+
"grad_norm": 0.39349165558815,
|
| 826 |
+
"learning_rate": 8.289225491142292e-05,
|
| 827 |
+
"loss": 0.0559,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.1052631578947367,
|
| 832 |
+
"grad_norm": 0.4374188482761383,
|
| 833 |
+
"learning_rate": 8.256329875734375e-05,
|
| 834 |
+
"loss": 0.0544,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.1315789473684212,
|
| 839 |
+
"grad_norm": 0.3512132465839386,
|
| 840 |
+
"learning_rate": 8.223187650149712e-05,
|
| 841 |
+
"loss": 0.0477,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.1578947368421053,
|
| 846 |
+
"grad_norm": 0.3832947313785553,
|
| 847 |
+
"learning_rate": 8.189801324333681e-05,
|
| 848 |
+
"loss": 0.0453,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.1842105263157894,
|
| 853 |
+
"grad_norm": 0.27593138813972473,
|
| 854 |
+
"learning_rate": 8.156173426717988e-05,
|
| 855 |
+
"loss": 0.0529,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.2105263157894735,
|
| 860 |
+
"grad_norm": 0.3951173424720764,
|
| 861 |
+
"learning_rate": 8.122306504029194e-05,
|
| 862 |
+
"loss": 0.0524,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.236842105263158,
|
| 867 |
+
"grad_norm": 0.3123633563518524,
|
| 868 |
+
"learning_rate": 8.08820312109583e-05,
|
| 869 |
+
"loss": 0.0477,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.263157894736842,
|
| 874 |
+
"grad_norm": 0.4703565835952759,
|
| 875 |
+
"learning_rate": 8.053865860654175e-05,
|
| 876 |
+
"loss": 0.0473,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 3.2894736842105265,
|
| 881 |
+
"grad_norm": 0.3525274991989136,
|
| 882 |
+
"learning_rate": 8.019297323152642e-05,
|
| 883 |
+
"loss": 0.057,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.3157894736842106,
|
| 888 |
+
"grad_norm": 0.352622926235199,
|
| 889 |
+
"learning_rate": 7.984500126554853e-05,
|
| 890 |
+
"loss": 0.0487,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 3.3421052631578947,
|
| 895 |
+
"grad_norm": 0.5331799387931824,
|
| 896 |
+
"learning_rate": 7.94947690614136e-05,
|
| 897 |
+
"loss": 0.0519,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 3.3684210526315788,
|
| 902 |
+
"grad_norm": 0.5408539772033691,
|
| 903 |
+
"learning_rate": 7.914230314310079e-05,
|
| 904 |
+
"loss": 0.0467,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 3.3947368421052633,
|
| 909 |
+
"grad_norm": 0.42730775475502014,
|
| 910 |
+
"learning_rate": 7.878763020375415e-05,
|
| 911 |
+
"loss": 0.0486,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 3.4210526315789473,
|
| 916 |
+
"grad_norm": 0.3785395920276642,
|
| 917 |
+
"learning_rate": 7.843077710366105e-05,
|
| 918 |
+
"loss": 0.0518,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 3.4473684210526314,
|
| 923 |
+
"grad_norm": 0.3049032986164093,
|
| 924 |
+
"learning_rate": 7.807177086821802e-05,
|
| 925 |
+
"loss": 0.0433,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 3.473684210526316,
|
| 930 |
+
"grad_norm": 0.3494362533092499,
|
| 931 |
+
"learning_rate": 7.771063868588399e-05,
|
| 932 |
+
"loss": 0.0453,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 3.5,
|
| 937 |
+
"grad_norm": 0.4008173942565918,
|
| 938 |
+
"learning_rate": 7.734740790612136e-05,
|
| 939 |
+
"loss": 0.0544,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 3.526315789473684,
|
| 944 |
+
"grad_norm": 0.5177074074745178,
|
| 945 |
+
"learning_rate": 7.698210603732454e-05,
|
| 946 |
+
"loss": 0.0433,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 3.5526315789473686,
|
| 951 |
+
"grad_norm": 0.42569538950920105,
|
| 952 |
+
"learning_rate": 7.661476074473695e-05,
|
| 953 |
+
"loss": 0.0428,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 3.5789473684210527,
|
| 958 |
+
"grad_norm": 0.36168932914733887,
|
| 959 |
+
"learning_rate": 7.624539984835557e-05,
|
| 960 |
+
"loss": 0.0466,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 3.6052631578947367,
|
| 965 |
+
"grad_norm": 0.4134399890899658,
|
| 966 |
+
"learning_rate": 7.587405132082433e-05,
|
| 967 |
+
"loss": 0.0468,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 3.6315789473684212,
|
| 972 |
+
"grad_norm": 0.2959583103656769,
|
| 973 |
+
"learning_rate": 7.550074328531545e-05,
|
| 974 |
+
"loss": 0.0477,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 3.6578947368421053,
|
| 979 |
+
"grad_norm": 0.4948980510234833,
|
| 980 |
+
"learning_rate": 7.512550401339971e-05,
|
| 981 |
+
"loss": 0.0453,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 3.6842105263157894,
|
| 986 |
+
"grad_norm": 0.5422468185424805,
|
| 987 |
+
"learning_rate": 7.47483619229054e-05,
|
| 988 |
+
"loss": 0.049,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 3.7105263157894735,
|
| 993 |
+
"grad_norm": 0.4651060402393341,
|
| 994 |
+
"learning_rate": 7.436934557576612e-05,
|
| 995 |
+
"loss": 0.0486,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 3.736842105263158,
|
| 1000 |
+
"grad_norm": 0.37345531582832336,
|
| 1001 |
+
"learning_rate": 7.39884836758576e-05,
|
| 1002 |
+
"loss": 0.0428,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 3.763157894736842,
|
| 1007 |
+
"grad_norm": 0.4845902621746063,
|
| 1008 |
+
"learning_rate": 7.360580506682414e-05,
|
| 1009 |
+
"loss": 0.0435,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 3.7894736842105265,
|
| 1014 |
+
"grad_norm": 0.38114190101623535,
|
| 1015 |
+
"learning_rate": 7.322133872989398e-05,
|
| 1016 |
+
"loss": 0.0384,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 3.8157894736842106,
|
| 1021 |
+
"grad_norm": 0.4120785892009735,
|
| 1022 |
+
"learning_rate": 7.283511378168458e-05,
|
| 1023 |
+
"loss": 0.0422,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 3.8421052631578947,
|
| 1028 |
+
"grad_norm": 0.4169304370880127,
|
| 1029 |
+
"learning_rate": 7.244715947199749e-05,
|
| 1030 |
+
"loss": 0.0413,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 3.8684210526315788,
|
| 1035 |
+
"grad_norm": 0.3038704991340637,
|
| 1036 |
+
"learning_rate": 7.20575051816033e-05,
|
| 1037 |
+
"loss": 0.0445,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 3.8947368421052633,
|
| 1042 |
+
"grad_norm": 0.3277107775211334,
|
| 1043 |
+
"learning_rate": 7.16661804200164e-05,
|
| 1044 |
+
"loss": 0.0443,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 3.9210526315789473,
|
| 1049 |
+
"grad_norm": 0.3675183057785034,
|
| 1050 |
+
"learning_rate": 7.127321482326026e-05,
|
| 1051 |
+
"loss": 0.046,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 3.9473684210526314,
|
| 1056 |
+
"grad_norm": 0.3455309569835663,
|
| 1057 |
+
"learning_rate": 7.087863815162298e-05,
|
| 1058 |
+
"loss": 0.0424,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 3.973684210526316,
|
| 1063 |
+
"grad_norm": 0.49531644582748413,
|
| 1064 |
+
"learning_rate": 7.04824802874035e-05,
|
| 1065 |
+
"loss": 0.0457,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.0,
|
| 1070 |
+
"grad_norm": 0.6025938391685486,
|
| 1071 |
+
"learning_rate": 7.008477123264848e-05,
|
| 1072 |
+
"loss": 0.0497,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.026315789473684,
|
| 1077 |
+
"grad_norm": 0.2605507969856262,
|
| 1078 |
+
"learning_rate": 6.96855411068802e-05,
|
| 1079 |
+
"loss": 0.047,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.052631578947368,
|
| 1084 |
+
"grad_norm": 0.3557581603527069,
|
| 1085 |
+
"learning_rate": 6.928482014481558e-05,
|
| 1086 |
+
"loss": 0.0434,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 4.078947368421052,
|
| 1091 |
+
"grad_norm": 0.6790323257446289,
|
| 1092 |
+
"learning_rate": 6.888263869407631e-05,
|
| 1093 |
+
"loss": 0.0431,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 4.105263157894737,
|
| 1098 |
+
"grad_norm": 0.5850968360900879,
|
| 1099 |
+
"learning_rate": 6.847902721289068e-05,
|
| 1100 |
+
"loss": 0.0444,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 4.131578947368421,
|
| 1105 |
+
"grad_norm": 0.22265614569187164,
|
| 1106 |
+
"learning_rate": 6.807401626778679e-05,
|
| 1107 |
+
"loss": 0.0463,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 4.157894736842105,
|
| 1112 |
+
"grad_norm": 0.3996434509754181,
|
| 1113 |
+
"learning_rate": 6.766763653127773e-05,
|
| 1114 |
+
"loss": 0.0391,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 4.184210526315789,
|
| 1119 |
+
"grad_norm": 0.4038737118244171,
|
| 1120 |
+
"learning_rate": 6.725991877953868e-05,
|
| 1121 |
+
"loss": 0.0432,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 4.2105263157894735,
|
| 1126 |
+
"grad_norm": 0.5624067187309265,
|
| 1127 |
+
"learning_rate": 6.685089389007612e-05,
|
| 1128 |
+
"loss": 0.0437,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 4.2368421052631575,
|
| 1133 |
+
"grad_norm": 0.32445138692855835,
|
| 1134 |
+
"learning_rate": 6.644059283938938e-05,
|
| 1135 |
+
"loss": 0.0439,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 4.2631578947368425,
|
| 1140 |
+
"grad_norm": 0.385633647441864,
|
| 1141 |
+
"learning_rate": 6.602904670062476e-05,
|
| 1142 |
+
"loss": 0.0362,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 4.2894736842105265,
|
| 1147 |
+
"grad_norm": 0.4192872643470764,
|
| 1148 |
+
"learning_rate": 6.561628664122226e-05,
|
| 1149 |
+
"loss": 0.0439,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 4.315789473684211,
|
| 1154 |
+
"grad_norm": 0.46780234575271606,
|
| 1155 |
+
"learning_rate": 6.520234392055522e-05,
|
| 1156 |
+
"loss": 0.0447,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 4.342105263157895,
|
| 1161 |
+
"grad_norm": 0.30417683720588684,
|
| 1162 |
+
"learning_rate": 6.478724988756285e-05,
|
| 1163 |
+
"loss": 0.0413,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 4.368421052631579,
|
| 1168 |
+
"grad_norm": 0.3585522472858429,
|
| 1169 |
+
"learning_rate": 6.437103597837631e-05,
|
| 1170 |
+
"loss": 0.0451,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 4.394736842105263,
|
| 1175 |
+
"grad_norm": 0.2543241083621979,
|
| 1176 |
+
"learning_rate": 6.39537337139377e-05,
|
| 1177 |
+
"loss": 0.0454,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 4.421052631578947,
|
| 1182 |
+
"grad_norm": 0.41101086139678955,
|
| 1183 |
+
"learning_rate": 6.353537469761315e-05,
|
| 1184 |
+
"loss": 0.0468,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 4.447368421052632,
|
| 1189 |
+
"grad_norm": 0.2859801650047302,
|
| 1190 |
+
"learning_rate": 6.311599061279932e-05,
|
| 1191 |
+
"loss": 0.0462,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 4.473684210526316,
|
| 1196 |
+
"grad_norm": 0.2935991585254669,
|
| 1197 |
+
"learning_rate": 6.269561322052378e-05,
|
| 1198 |
+
"loss": 0.0457,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 4.5,
|
| 1203 |
+
"grad_norm": 0.311879426240921,
|
| 1204 |
+
"learning_rate": 6.227427435703997e-05,
|
| 1205 |
+
"loss": 0.0419,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 4.526315789473684,
|
| 1210 |
+
"grad_norm": 0.4428638517856598,
|
| 1211 |
+
"learning_rate": 6.185200593141593e-05,
|
| 1212 |
+
"loss": 0.0412,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 4.552631578947368,
|
| 1217 |
+
"grad_norm": 0.44332432746887207,
|
| 1218 |
+
"learning_rate": 6.142883992311781e-05,
|
| 1219 |
+
"loss": 0.0391,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 4.578947368421053,
|
| 1224 |
+
"grad_norm": 0.31303486227989197,
|
| 1225 |
+
"learning_rate": 6.100480837958802e-05,
|
| 1226 |
+
"loss": 0.0478,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 4.605263157894737,
|
| 1231 |
+
"grad_norm": 0.4248807430267334,
|
| 1232 |
+
"learning_rate": 6.057994341381813e-05,
|
| 1233 |
+
"loss": 0.0419,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 4.631578947368421,
|
| 1238 |
+
"grad_norm": 0.38378843665122986,
|
| 1239 |
+
"learning_rate": 6.015427720191693e-05,
|
| 1240 |
+
"loss": 0.0388,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 4.657894736842105,
|
| 1245 |
+
"grad_norm": 0.232202410697937,
|
| 1246 |
+
"learning_rate": 5.9727841980673604e-05,
|
| 1247 |
+
"loss": 0.0491,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 4.684210526315789,
|
| 1252 |
+
"grad_norm": 0.3846134841442108,
|
| 1253 |
+
"learning_rate": 5.93006700451164e-05,
|
| 1254 |
+
"loss": 0.0442,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 4.7105263157894735,
|
| 1259 |
+
"grad_norm": 0.24118691682815552,
|
| 1260 |
+
"learning_rate": 5.887279374606679e-05,
|
| 1261 |
+
"loss": 0.0379,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 4.7368421052631575,
|
| 1266 |
+
"grad_norm": 0.3538924753665924,
|
| 1267 |
+
"learning_rate": 5.844424548768952e-05,
|
| 1268 |
+
"loss": 0.0371,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 4.7631578947368425,
|
| 1273 |
+
"grad_norm": 0.2391858547925949,
|
| 1274 |
+
"learning_rate": 5.8015057725038534e-05,
|
| 1275 |
+
"loss": 0.0393,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 4.7894736842105265,
|
| 1280 |
+
"grad_norm": 0.43974366784095764,
|
| 1281 |
+
"learning_rate": 5.7585262961599054e-05,
|
| 1282 |
+
"loss": 0.0408,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 4.815789473684211,
|
| 1287 |
+
"grad_norm": 0.3488527834415436,
|
| 1288 |
+
"learning_rate": 5.7154893746826014e-05,
|
| 1289 |
+
"loss": 0.0419,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 4.842105263157895,
|
| 1294 |
+
"grad_norm": 0.5002457499504089,
|
| 1295 |
+
"learning_rate": 5.672398267367902e-05,
|
| 1296 |
+
"loss": 0.0403,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 4.868421052631579,
|
| 1301 |
+
"grad_norm": 0.31320708990097046,
|
| 1302 |
+
"learning_rate": 5.6292562376154037e-05,
|
| 1303 |
+
"loss": 0.0423,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 4.894736842105263,
|
| 1308 |
+
"grad_norm": 0.26685863733291626,
|
| 1309 |
+
"learning_rate": 5.586066552681179e-05,
|
| 1310 |
+
"loss": 0.0377,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 4.921052631578947,
|
| 1315 |
+
"grad_norm": 0.3947713077068329,
|
| 1316 |
+
"learning_rate": 5.542832483430363e-05,
|
| 1317 |
+
"loss": 0.0401,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 4.947368421052632,
|
| 1322 |
+
"grad_norm": 0.20824529230594635,
|
| 1323 |
+
"learning_rate": 5.499557304089419e-05,
|
| 1324 |
+
"loss": 0.0382,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 4.973684210526316,
|
| 1329 |
+
"grad_norm": 0.25169357657432556,
|
| 1330 |
+
"learning_rate": 5.4562442919981816e-05,
|
| 1331 |
+
"loss": 0.0339,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 5.0,
|
| 1336 |
+
"grad_norm": 0.4444282054901123,
|
| 1337 |
+
"learning_rate": 5.4128967273616625e-05,
|
| 1338 |
+
"loss": 0.0443,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 5.026315789473684,
|
| 1343 |
+
"grad_norm": 0.4260099530220032,
|
| 1344 |
+
"learning_rate": 5.3695178930016196e-05,
|
| 1345 |
+
"loss": 0.0364,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 5.052631578947368,
|
| 1350 |
+
"grad_norm": 0.33936482667922974,
|
| 1351 |
+
"learning_rate": 5.3261110741079525e-05,
|
| 1352 |
+
"loss": 0.0367,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 5.078947368421052,
|
| 1357 |
+
"grad_norm": 0.5758324861526489,
|
| 1358 |
+
"learning_rate": 5.2826795579898956e-05,
|
| 1359 |
+
"loss": 0.0383,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 5.105263157894737,
|
| 1364 |
+
"grad_norm": 0.4980020821094513,
|
| 1365 |
+
"learning_rate": 5.2392266338270736e-05,
|
| 1366 |
+
"loss": 0.0422,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 5.131578947368421,
|
| 1371 |
+
"grad_norm": 0.4701398015022278,
|
| 1372 |
+
"learning_rate": 5.195755592420387e-05,
|
| 1373 |
+
"loss": 0.0372,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 5.157894736842105,
|
| 1378 |
+
"grad_norm": 0.3783451020717621,
|
| 1379 |
+
"learning_rate": 5.1522697259428146e-05,
|
| 1380 |
+
"loss": 0.0371,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 5.184210526315789,
|
| 1385 |
+
"grad_norm": 0.41548261046409607,
|
| 1386 |
+
"learning_rate": 5.1087723276900646e-05,
|
| 1387 |
+
"loss": 0.0415,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 5.2105263157894735,
|
| 1392 |
+
"grad_norm": 0.3551521897315979,
|
| 1393 |
+
"learning_rate": 5.065266691831181e-05,
|
| 1394 |
+
"loss": 0.0344,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 5.2368421052631575,
|
| 1399 |
+
"grad_norm": 0.5270403623580933,
|
| 1400 |
+
"learning_rate": 5.021756113159062e-05,
|
| 1401 |
+
"loss": 0.0355,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 5.2631578947368425,
|
| 1406 |
+
"grad_norm": 0.29675352573394775,
|
| 1407 |
+
"learning_rate": 4.978243886840939e-05,
|
| 1408 |
+
"loss": 0.0424,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 5.2894736842105265,
|
| 1413 |
+
"grad_norm": 0.3667472302913666,
|
| 1414 |
+
"learning_rate": 4.934733308168821e-05,
|
| 1415 |
+
"loss": 0.0449,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 5.315789473684211,
|
| 1420 |
+
"grad_norm": 0.34768638014793396,
|
| 1421 |
+
"learning_rate": 4.891227672309935e-05,
|
| 1422 |
+
"loss": 0.041,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 5.342105263157895,
|
| 1427 |
+
"grad_norm": 0.38154706358909607,
|
| 1428 |
+
"learning_rate": 4.8477302740571866e-05,
|
| 1429 |
+
"loss": 0.0382,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 5.368421052631579,
|
| 1434 |
+
"grad_norm": 0.16855105757713318,
|
| 1435 |
+
"learning_rate": 4.804244407579613e-05,
|
| 1436 |
+
"loss": 0.0376,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 5.394736842105263,
|
| 1441 |
+
"grad_norm": 0.3421989679336548,
|
| 1442 |
+
"learning_rate": 4.760773366172929e-05,
|
| 1443 |
+
"loss": 0.0381,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 5.421052631578947,
|
| 1448 |
+
"grad_norm": 0.3012019991874695,
|
| 1449 |
+
"learning_rate": 4.717320442010105e-05,
|
| 1450 |
+
"loss": 0.0353,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 5.447368421052632,
|
| 1455 |
+
"grad_norm": 0.2907329797744751,
|
| 1456 |
+
"learning_rate": 4.673888925892048e-05,
|
| 1457 |
+
"loss": 0.0372,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 5.473684210526316,
|
| 1462 |
+
"grad_norm": 0.3655814826488495,
|
| 1463 |
+
"learning_rate": 4.630482106998381e-05,
|
| 1464 |
+
"loss": 0.0354,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 5.5,
|
| 1469 |
+
"grad_norm": 0.3697777986526489,
|
| 1470 |
+
"learning_rate": 4.5871032726383386e-05,
|
| 1471 |
+
"loss": 0.0426,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 5.526315789473684,
|
| 1476 |
+
"grad_norm": 0.22509945929050446,
|
| 1477 |
+
"learning_rate": 4.5437557080018175e-05,
|
| 1478 |
+
"loss": 0.0358,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 5.552631578947368,
|
| 1483 |
+
"grad_norm": 0.30516788363456726,
|
| 1484 |
+
"learning_rate": 4.500442695910582e-05,
|
| 1485 |
+
"loss": 0.0411,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 5.578947368421053,
|
| 1490 |
+
"grad_norm": 0.3428977429866791,
|
| 1491 |
+
"learning_rate": 4.457167516569637e-05,
|
| 1492 |
+
"loss": 0.032,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 5.605263157894737,
|
| 1497 |
+
"grad_norm": 0.2875311076641083,
|
| 1498 |
+
"learning_rate": 4.413933447318821e-05,
|
| 1499 |
+
"loss": 0.0339,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 5.631578947368421,
|
| 1504 |
+
"grad_norm": 0.3058890700340271,
|
| 1505 |
+
"learning_rate": 4.3707437623845995e-05,
|
| 1506 |
+
"loss": 0.035,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 5.657894736842105,
|
| 1511 |
+
"grad_norm": 0.317531019449234,
|
| 1512 |
+
"learning_rate": 4.3276017326320985e-05,
|
| 1513 |
+
"loss": 0.0324,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 5.684210526315789,
|
| 1518 |
+
"grad_norm": 0.3118530809879303,
|
| 1519 |
+
"learning_rate": 4.2845106253174e-05,
|
| 1520 |
+
"loss": 0.0345,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 5.7105263157894735,
|
| 1525 |
+
"grad_norm": 0.2496698647737503,
|
| 1526 |
+
"learning_rate": 4.2414737038400964e-05,
|
| 1527 |
+
"loss": 0.0373,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 5.7368421052631575,
|
| 1532 |
+
"grad_norm": 0.26971638202667236,
|
| 1533 |
+
"learning_rate": 4.198494227496148e-05,
|
| 1534 |
+
"loss": 0.0368,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 5.7631578947368425,
|
| 1539 |
+
"grad_norm": 0.43503808975219727,
|
| 1540 |
+
"learning_rate": 4.155575451231048e-05,
|
| 1541 |
+
"loss": 0.0348,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 5.7894736842105265,
|
| 1546 |
+
"grad_norm": 0.26233163475990295,
|
| 1547 |
+
"learning_rate": 4.112720625393322e-05,
|
| 1548 |
+
"loss": 0.0354,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 5.815789473684211,
|
| 1553 |
+
"grad_norm": 0.40549346804618835,
|
| 1554 |
+
"learning_rate": 4.069932995488361e-05,
|
| 1555 |
+
"loss": 0.0385,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 5.842105263157895,
|
| 1560 |
+
"grad_norm": 0.3487369120121002,
|
| 1561 |
+
"learning_rate": 4.0272158019326414e-05,
|
| 1562 |
+
"loss": 0.0365,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 5.868421052631579,
|
| 1567 |
+
"grad_norm": 0.35258445143699646,
|
| 1568 |
+
"learning_rate": 3.9845722798083066e-05,
|
| 1569 |
+
"loss": 0.035,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 5.894736842105263,
|
| 1574 |
+
"grad_norm": 0.3888448476791382,
|
| 1575 |
+
"learning_rate": 3.942005658618188e-05,
|
| 1576 |
+
"loss": 0.0357,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 5.921052631578947,
|
| 1581 |
+
"grad_norm": 0.41619524359703064,
|
| 1582 |
+
"learning_rate": 3.8995191620412e-05,
|
| 1583 |
+
"loss": 0.0352,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 5.947368421052632,
|
| 1588 |
+
"grad_norm": 0.5103023052215576,
|
| 1589 |
+
"learning_rate": 3.8571160076882204e-05,
|
| 1590 |
+
"loss": 0.0331,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 5.973684210526316,
|
| 1595 |
+
"grad_norm": 0.458965927362442,
|
| 1596 |
+
"learning_rate": 3.8147994068584087e-05,
|
| 1597 |
+
"loss": 0.0301,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 6.0,
|
| 1602 |
+
"grad_norm": 0.32402315735816956,
|
| 1603 |
+
"learning_rate": 3.772572564296005e-05,
|
| 1604 |
+
"loss": 0.0351,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 6.026315789473684,
|
| 1609 |
+
"grad_norm": 0.2968176603317261,
|
| 1610 |
+
"learning_rate": 3.730438677947624e-05,
|
| 1611 |
+
"loss": 0.0413,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 6.052631578947368,
|
| 1616 |
+
"grad_norm": 0.3646889626979828,
|
| 1617 |
+
"learning_rate": 3.6884009387200714e-05,
|
| 1618 |
+
"loss": 0.0357,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 6.078947368421052,
|
| 1623 |
+
"grad_norm": 0.26003819704055786,
|
| 1624 |
+
"learning_rate": 3.646462530238684e-05,
|
| 1625 |
+
"loss": 0.0334,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 6.105263157894737,
|
| 1630 |
+
"grad_norm": 0.28209343552589417,
|
| 1631 |
+
"learning_rate": 3.60462662860623e-05,
|
| 1632 |
+
"loss": 0.0317,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 6.131578947368421,
|
| 1637 |
+
"grad_norm": 0.4188937544822693,
|
| 1638 |
+
"learning_rate": 3.56289640216237e-05,
|
| 1639 |
+
"loss": 0.0354,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 6.157894736842105,
|
| 1644 |
+
"grad_norm": 0.44584834575653076,
|
| 1645 |
+
"learning_rate": 3.521275011243715e-05,
|
| 1646 |
+
"loss": 0.0395,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 6.184210526315789,
|
| 1651 |
+
"grad_norm": 0.44341957569122314,
|
| 1652 |
+
"learning_rate": 3.4797656079444806e-05,
|
| 1653 |
+
"loss": 0.033,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 6.2105263157894735,
|
| 1658 |
+
"grad_norm": 0.3678707182407379,
|
| 1659 |
+
"learning_rate": 3.4383713358777735e-05,
|
| 1660 |
+
"loss": 0.0344,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 6.2368421052631575,
|
| 1665 |
+
"grad_norm": 0.30562394857406616,
|
| 1666 |
+
"learning_rate": 3.397095329937526e-05,
|
| 1667 |
+
"loss": 0.0339,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 6.2631578947368425,
|
| 1672 |
+
"grad_norm": 0.4624996781349182,
|
| 1673 |
+
"learning_rate": 3.355940716061064e-05,
|
| 1674 |
+
"loss": 0.0341,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 6.2894736842105265,
|
| 1679 |
+
"grad_norm": 0.20757588744163513,
|
| 1680 |
+
"learning_rate": 3.31491061099239e-05,
|
| 1681 |
+
"loss": 0.0296,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 6.315789473684211,
|
| 1686 |
+
"grad_norm": 0.31677284836769104,
|
| 1687 |
+
"learning_rate": 3.274008122046132e-05,
|
| 1688 |
+
"loss": 0.0338,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 6.342105263157895,
|
| 1693 |
+
"grad_norm": 0.24551235139369965,
|
| 1694 |
+
"learning_rate": 3.233236346872227e-05,
|
| 1695 |
+
"loss": 0.0381,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 6.368421052631579,
|
| 1700 |
+
"grad_norm": 0.250813364982605,
|
| 1701 |
+
"learning_rate": 3.192598373221322e-05,
|
| 1702 |
+
"loss": 0.0333,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 6.394736842105263,
|
| 1707 |
+
"grad_norm": 0.5114396810531616,
|
| 1708 |
+
"learning_rate": 3.152097278710933e-05,
|
| 1709 |
+
"loss": 0.0396,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 6.421052631578947,
|
| 1714 |
+
"grad_norm": 0.4336398243904114,
|
| 1715 |
+
"learning_rate": 3.1117361305923684e-05,
|
| 1716 |
+
"loss": 0.0361,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 6.447368421052632,
|
| 1721 |
+
"grad_norm": 0.3308594226837158,
|
| 1722 |
+
"learning_rate": 3.071517985518442e-05,
|
| 1723 |
+
"loss": 0.0314,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 6.473684210526316,
|
| 1728 |
+
"grad_norm": 0.4118829667568207,
|
| 1729 |
+
"learning_rate": 3.0314458893119808e-05,
|
| 1730 |
+
"loss": 0.0337,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 6.5,
|
| 1735 |
+
"grad_norm": 0.19590795040130615,
|
| 1736 |
+
"learning_rate": 2.991522876735154e-05,
|
| 1737 |
+
"loss": 0.0359,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 6.526315789473684,
|
| 1742 |
+
"grad_norm": 0.39492878317832947,
|
| 1743 |
+
"learning_rate": 2.9517519712596498e-05,
|
| 1744 |
+
"loss": 0.0304,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 6.552631578947368,
|
| 1749 |
+
"grad_norm": 0.2706056833267212,
|
| 1750 |
+
"learning_rate": 2.9121361848377014e-05,
|
| 1751 |
+
"loss": 0.0301,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 6.578947368421053,
|
| 1756 |
+
"grad_norm": 0.2509630024433136,
|
| 1757 |
+
"learning_rate": 2.872678517673975e-05,
|
| 1758 |
+
"loss": 0.0313,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 6.605263157894737,
|
| 1763 |
+
"grad_norm": 0.37828516960144043,
|
| 1764 |
+
"learning_rate": 2.8333819579983623e-05,
|
| 1765 |
+
"loss": 0.0323,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 6.631578947368421,
|
| 1770 |
+
"grad_norm": 0.3628530204296112,
|
| 1771 |
+
"learning_rate": 2.794249481839669e-05,
|
| 1772 |
+
"loss": 0.0292,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 6.657894736842105,
|
| 1777 |
+
"grad_norm": 0.35174816846847534,
|
| 1778 |
+
"learning_rate": 2.7552840528002498e-05,
|
| 1779 |
+
"loss": 0.0326,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 6.684210526315789,
|
| 1784 |
+
"grad_norm": 0.3942660391330719,
|
| 1785 |
+
"learning_rate": 2.7164886218315444e-05,
|
| 1786 |
+
"loss": 0.0314,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 6.7105263157894735,
|
| 1791 |
+
"grad_norm": 0.2891586124897003,
|
| 1792 |
+
"learning_rate": 2.6778661270106025e-05,
|
| 1793 |
+
"loss": 0.0346,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 6.7368421052631575,
|
| 1798 |
+
"grad_norm": 0.3602871000766754,
|
| 1799 |
+
"learning_rate": 2.6394194933175875e-05,
|
| 1800 |
+
"loss": 0.0338,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 6.7631578947368425,
|
| 1805 |
+
"grad_norm": 0.3285452127456665,
|
| 1806 |
+
"learning_rate": 2.601151632414241e-05,
|
| 1807 |
+
"loss": 0.0301,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 6.7894736842105265,
|
| 1812 |
+
"grad_norm": 0.3526369035243988,
|
| 1813 |
+
"learning_rate": 2.5630654424233903e-05,
|
| 1814 |
+
"loss": 0.0331,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 6.815789473684211,
|
| 1819 |
+
"grad_norm": 0.2297569215297699,
|
| 1820 |
+
"learning_rate": 2.5251638077094602e-05,
|
| 1821 |
+
"loss": 0.0311,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 6.842105263157895,
|
| 1826 |
+
"grad_norm": 0.24023893475532532,
|
| 1827 |
+
"learning_rate": 2.4874495986600294e-05,
|
| 1828 |
+
"loss": 0.0299,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 6.868421052631579,
|
| 1833 |
+
"grad_norm": 0.3183501660823822,
|
| 1834 |
+
"learning_rate": 2.4499256714684565e-05,
|
| 1835 |
+
"loss": 0.0329,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 6.894736842105263,
|
| 1840 |
+
"grad_norm": 0.335347056388855,
|
| 1841 |
+
"learning_rate": 2.4125948679175686e-05,
|
| 1842 |
+
"loss": 0.0302,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 6.921052631578947,
|
| 1847 |
+
"grad_norm": 0.31279057264328003,
|
| 1848 |
+
"learning_rate": 2.3754600151644445e-05,
|
| 1849 |
+
"loss": 0.0317,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 6.947368421052632,
|
| 1854 |
+
"grad_norm": 0.25722017884254456,
|
| 1855 |
+
"learning_rate": 2.3385239255263077e-05,
|
| 1856 |
+
"loss": 0.034,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 6.973684210526316,
|
| 1861 |
+
"grad_norm": 0.4903739392757416,
|
| 1862 |
+
"learning_rate": 2.3017893962675458e-05,
|
| 1863 |
+
"loss": 0.0272,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 7.0,
|
| 1868 |
+
"grad_norm": 0.31394830346107483,
|
| 1869 |
+
"learning_rate": 2.2652592093878666e-05,
|
| 1870 |
+
"loss": 0.0343,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 7.026315789473684,
|
| 1875 |
+
"grad_norm": 0.24519610404968262,
|
| 1876 |
+
"learning_rate": 2.228936131411601e-05,
|
| 1877 |
+
"loss": 0.0283,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 7.052631578947368,
|
| 1882 |
+
"grad_norm": 0.3303487002849579,
|
| 1883 |
+
"learning_rate": 2.1928229131782007e-05,
|
| 1884 |
+
"loss": 0.0311,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 7.078947368421052,
|
| 1889 |
+
"grad_norm": 0.18910463154315948,
|
| 1890 |
+
"learning_rate": 2.1569222896338966e-05,
|
| 1891 |
+
"loss": 0.0272,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 7.105263157894737,
|
| 1896 |
+
"grad_norm": 0.2918621003627777,
|
| 1897 |
+
"learning_rate": 2.1212369796245864e-05,
|
| 1898 |
+
"loss": 0.0298,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 7.131578947368421,
|
| 1903 |
+
"grad_norm": 0.42113614082336426,
|
| 1904 |
+
"learning_rate": 2.0857696856899232e-05,
|
| 1905 |
+
"loss": 0.0295,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 7.157894736842105,
|
| 1910 |
+
"grad_norm": 0.3034079670906067,
|
| 1911 |
+
"learning_rate": 2.0505230938586418e-05,
|
| 1912 |
+
"loss": 0.0235,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 7.184210526315789,
|
| 1917 |
+
"grad_norm": 0.4173114001750946,
|
| 1918 |
+
"learning_rate": 2.0154998734451474e-05,
|
| 1919 |
+
"loss": 0.0302,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 7.2105263157894735,
|
| 1924 |
+
"grad_norm": 0.3530351519584656,
|
| 1925 |
+
"learning_rate": 1.980702676847358e-05,
|
| 1926 |
+
"loss": 0.0322,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 7.2368421052631575,
|
| 1931 |
+
"grad_norm": 0.29570046067237854,
|
| 1932 |
+
"learning_rate": 1.9461341393458254e-05,
|
| 1933 |
+
"loss": 0.0272,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 7.2631578947368425,
|
| 1938 |
+
"grad_norm": 0.2424498051404953,
|
| 1939 |
+
"learning_rate": 1.9117968789041712e-05,
|
| 1940 |
+
"loss": 0.028,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 7.2894736842105265,
|
| 1945 |
+
"grad_norm": 0.3934868574142456,
|
| 1946 |
+
"learning_rate": 1.877693495970809e-05,
|
| 1947 |
+
"loss": 0.0272,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 7.315789473684211,
|
| 1952 |
+
"grad_norm": 0.21242979168891907,
|
| 1953 |
+
"learning_rate": 1.8438265732820126e-05,
|
| 1954 |
+
"loss": 0.0327,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 7.342105263157895,
|
| 1959 |
+
"grad_norm": 0.37688693404197693,
|
| 1960 |
+
"learning_rate": 1.8101986756663197e-05,
|
| 1961 |
+
"loss": 0.0351,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 7.368421052631579,
|
| 1966 |
+
"grad_norm": 0.22811609506607056,
|
| 1967 |
+
"learning_rate": 1.776812349850289e-05,
|
| 1968 |
+
"loss": 0.0276,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 7.394736842105263,
|
| 1973 |
+
"grad_norm": 0.22656874358654022,
|
| 1974 |
+
"learning_rate": 1.7436701242656272e-05,
|
| 1975 |
+
"loss": 0.0287,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 7.421052631578947,
|
| 1980 |
+
"grad_norm": 0.35913366079330444,
|
| 1981 |
+
"learning_rate": 1.7107745088577087e-05,
|
| 1982 |
+
"loss": 0.0309,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 7.447368421052632,
|
| 1987 |
+
"grad_norm": 0.23046566545963287,
|
| 1988 |
+
"learning_rate": 1.678127994895492e-05,
|
| 1989 |
+
"loss": 0.0242,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 7.473684210526316,
|
| 1994 |
+
"grad_norm": 0.5616809725761414,
|
| 1995 |
+
"learning_rate": 1.6457330547828402e-05,
|
| 1996 |
+
"loss": 0.0292,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 7.5,
|
| 2001 |
+
"grad_norm": 0.25659796595573425,
|
| 2002 |
+
"learning_rate": 1.6135921418712956e-05,
|
| 2003 |
+
"loss": 0.0265,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 7.526315789473684,
|
| 2008 |
+
"grad_norm": 0.32121288776397705,
|
| 2009 |
+
"learning_rate": 1.5817076902742622e-05,
|
| 2010 |
+
"loss": 0.0283,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 7.552631578947368,
|
| 2015 |
+
"grad_norm": 0.4093388319015503,
|
| 2016 |
+
"learning_rate": 1.5500821146826805e-05,
|
| 2017 |
+
"loss": 0.0338,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 7.578947368421053,
|
| 2022 |
+
"grad_norm": 0.4066798686981201,
|
| 2023 |
+
"learning_rate": 1.5187178101821503e-05,
|
| 2024 |
+
"loss": 0.0354,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 7.605263157894737,
|
| 2029 |
+
"grad_norm": 0.2548167109489441,
|
| 2030 |
+
"learning_rate": 1.4876171520715399e-05,
|
| 2031 |
+
"loss": 0.0289,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 7.631578947368421,
|
| 2036 |
+
"grad_norm": 0.375201940536499,
|
| 2037 |
+
"learning_rate": 1.4567824956831043e-05,
|
| 2038 |
+
"loss": 0.0308,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 7.657894736842105,
|
| 2043 |
+
"grad_norm": 0.2906716763973236,
|
| 2044 |
+
"learning_rate": 1.4262161762041121e-05,
|
| 2045 |
+
"loss": 0.03,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 7.684210526315789,
|
| 2050 |
+
"grad_norm": 0.25092822313308716,
|
| 2051 |
+
"learning_rate": 1.3959205084999911e-05,
|
| 2052 |
+
"loss": 0.0262,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 7.7105263157894735,
|
| 2057 |
+
"grad_norm": 0.27973148226737976,
|
| 2058 |
+
"learning_rate": 1.3658977869390166e-05,
|
| 2059 |
+
"loss": 0.0259,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 7.7368421052631575,
|
| 2064 |
+
"grad_norm": 0.24641232192516327,
|
| 2065 |
+
"learning_rate": 1.336150285218558e-05,
|
| 2066 |
+
"loss": 0.0288,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 7.7631578947368425,
|
| 2071 |
+
"grad_norm": 0.17812731862068176,
|
| 2072 |
+
"learning_rate": 1.3066802561928854e-05,
|
| 2073 |
+
"loss": 0.0302,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 7.7894736842105265,
|
| 2078 |
+
"grad_norm": 0.2735963761806488,
|
| 2079 |
+
"learning_rate": 1.2774899317025468e-05,
|
| 2080 |
+
"loss": 0.0281,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 7.815789473684211,
|
| 2085 |
+
"grad_norm": 0.2304324209690094,
|
| 2086 |
+
"learning_rate": 1.2485815224053582e-05,
|
| 2087 |
+
"loss": 0.0286,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 7.842105263157895,
|
| 2092 |
+
"grad_norm": 0.28357961773872375,
|
| 2093 |
+
"learning_rate": 1.2199572176089741e-05,
|
| 2094 |
+
"loss": 0.0228,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 7.868421052631579,
|
| 2099 |
+
"grad_norm": 0.20956674218177795,
|
| 2100 |
+
"learning_rate": 1.1916191851050873e-05,
|
| 2101 |
+
"loss": 0.0276,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 7.894736842105263,
|
| 2106 |
+
"grad_norm": 0.25499334931373596,
|
| 2107 |
+
"learning_rate": 1.163569571005269e-05,
|
| 2108 |
+
"loss": 0.0263,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 7.921052631578947,
|
| 2113 |
+
"grad_norm": 0.406367689371109,
|
| 2114 |
+
"learning_rate": 1.1358104995784186e-05,
|
| 2115 |
+
"loss": 0.03,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 7.947368421052632,
|
| 2120 |
+
"grad_norm": 0.49673014879226685,
|
| 2121 |
+
"learning_rate": 1.1083440730898974e-05,
|
| 2122 |
+
"loss": 0.029,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 7.973684210526316,
|
| 2127 |
+
"grad_norm": 0.21329355239868164,
|
| 2128 |
+
"learning_rate": 1.0811723716423233e-05,
|
| 2129 |
+
"loss": 0.0254,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 8.0,
|
| 2134 |
+
"grad_norm": 0.38303622603416443,
|
| 2135 |
+
"learning_rate": 1.0542974530180327e-05,
|
| 2136 |
+
"loss": 0.031,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 8.026315789473685,
|
| 2141 |
+
"grad_norm": 0.30326029658317566,
|
| 2142 |
+
"learning_rate": 1.027721352523237e-05,
|
| 2143 |
+
"loss": 0.031,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 8.052631578947368,
|
| 2148 |
+
"grad_norm": 0.18923179805278778,
|
| 2149 |
+
"learning_rate": 1.0014460828338928e-05,
|
| 2150 |
+
"loss": 0.0239,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 8.078947368421053,
|
| 2155 |
+
"grad_norm": 0.2916359603404999,
|
| 2156 |
+
"learning_rate": 9.75473633843268e-06,
|
| 2157 |
+
"loss": 0.0288,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 8.105263157894736,
|
| 2162 |
+
"grad_norm": 0.39194855093955994,
|
| 2163 |
+
"learning_rate": 9.498059725112467e-06,
|
| 2164 |
+
"loss": 0.0255,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 8.131578947368421,
|
| 2169 |
+
"grad_norm": 0.23311467468738556,
|
| 2170 |
+
"learning_rate": 9.244450427153683e-06,
|
| 2171 |
+
"loss": 0.0271,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 8.157894736842104,
|
| 2176 |
+
"grad_norm": 0.25806355476379395,
|
| 2177 |
+
"learning_rate": 8.99392765103605e-06,
|
| 2178 |
+
"loss": 0.0253,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 8.18421052631579,
|
| 2183 |
+
"grad_norm": 0.19991613924503326,
|
| 2184 |
+
"learning_rate": 8.746510369489103e-06,
|
| 2185 |
+
"loss": 0.0306,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 8.210526315789474,
|
| 2190 |
+
"grad_norm": 0.26584678888320923,
|
| 2191 |
+
"learning_rate": 8.502217320055427e-06,
|
| 2192 |
+
"loss": 0.0295,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 8.236842105263158,
|
| 2197 |
+
"grad_norm": 0.28366735577583313,
|
| 2198 |
+
"learning_rate": 8.261067003671447e-06,
|
| 2199 |
+
"loss": 0.0267,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 8.263157894736842,
|
| 2204 |
+
"grad_norm": 0.2124335616827011,
|
| 2205 |
+
"learning_rate": 8.0230776832664e-06,
|
| 2206 |
+
"loss": 0.027,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 8.289473684210526,
|
| 2211 |
+
"grad_norm": 0.6940687894821167,
|
| 2212 |
+
"learning_rate": 7.78826738237926e-06,
|
| 2213 |
+
"loss": 0.0239,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 8.31578947368421,
|
| 2218 |
+
"grad_norm": 0.3029802739620209,
|
| 2219 |
+
"learning_rate": 7.556653883793724e-06,
|
| 2220 |
+
"loss": 0.0293,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 8.342105263157896,
|
| 2225 |
+
"grad_norm": 0.33494675159454346,
|
| 2226 |
+
"learning_rate": 7.328254728191464e-06,
|
| 2227 |
+
"loss": 0.0322,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 8.368421052631579,
|
| 2232 |
+
"grad_norm": 0.5959463715553284,
|
| 2233 |
+
"learning_rate": 7.103087212823778e-06,
|
| 2234 |
+
"loss": 0.0282,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 8.394736842105264,
|
| 2239 |
+
"grad_norm": 0.3299950957298279,
|
| 2240 |
+
"learning_rate": 6.881168390201581e-06,
|
| 2241 |
+
"loss": 0.0289,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 8.421052631578947,
|
| 2246 |
+
"grad_norm": 0.28407934308052063,
|
| 2247 |
+
"learning_rate": 6.66251506680397e-06,
|
| 2248 |
+
"loss": 0.0277,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 8.447368421052632,
|
| 2253 |
+
"grad_norm": 0.24567903578281403,
|
| 2254 |
+
"learning_rate": 6.447143801805516e-06,
|
| 2255 |
+
"loss": 0.0293,
|
| 2256 |
+
"step": 3210
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 8.473684210526315,
|
| 2260 |
+
"grad_norm": 0.20823001861572266,
|
| 2261 |
+
"learning_rate": 6.23507090582206e-06,
|
| 2262 |
+
"loss": 0.0257,
|
| 2263 |
+
"step": 3220
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 8.5,
|
| 2267 |
+
"grad_norm": 0.29089733958244324,
|
| 2268 |
+
"learning_rate": 6.026312439675552e-06,
|
| 2269 |
+
"loss": 0.0268,
|
| 2270 |
+
"step": 3230
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 8.526315789473685,
|
| 2274 |
+
"grad_norm": 0.3374447524547577,
|
| 2275 |
+
"learning_rate": 5.820884213177713e-06,
|
| 2276 |
+
"loss": 0.0276,
|
| 2277 |
+
"step": 3240
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 8.552631578947368,
|
| 2281 |
+
"grad_norm": 0.32109740376472473,
|
| 2282 |
+
"learning_rate": 5.618801783932725e-06,
|
| 2283 |
+
"loss": 0.0302,
|
| 2284 |
+
"step": 3250
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 8.578947368421053,
|
| 2288 |
+
"grad_norm": 0.25355905294418335,
|
| 2289 |
+
"learning_rate": 5.420080456158971e-06,
|
| 2290 |
+
"loss": 0.0263,
|
| 2291 |
+
"step": 3260
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 8.605263157894736,
|
| 2295 |
+
"grad_norm": 0.6211085319519043,
|
| 2296 |
+
"learning_rate": 5.224735279530063e-06,
|
| 2297 |
+
"loss": 0.0284,
|
| 2298 |
+
"step": 3270
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 8.631578947368421,
|
| 2302 |
+
"grad_norm": 0.23430858552455902,
|
| 2303 |
+
"learning_rate": 5.032781048035034e-06,
|
| 2304 |
+
"loss": 0.0264,
|
| 2305 |
+
"step": 3280
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 8.657894736842106,
|
| 2309 |
+
"grad_norm": 0.24149766564369202,
|
| 2310 |
+
"learning_rate": 4.84423229885802e-06,
|
| 2311 |
+
"loss": 0.027,
|
| 2312 |
+
"step": 3290
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 8.68421052631579,
|
| 2316 |
+
"grad_norm": 0.3725610077381134,
|
| 2317 |
+
"learning_rate": 4.659103311277274e-06,
|
| 2318 |
+
"loss": 0.0253,
|
| 2319 |
+
"step": 3300
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 8.710526315789474,
|
| 2323 |
+
"grad_norm": 0.2504403293132782,
|
| 2324 |
+
"learning_rate": 4.477408105583741e-06,
|
| 2325 |
+
"loss": 0.0276,
|
| 2326 |
+
"step": 3310
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 8.736842105263158,
|
| 2330 |
+
"grad_norm": 0.21490493416786194,
|
| 2331 |
+
"learning_rate": 4.29916044201934e-06,
|
| 2332 |
+
"loss": 0.0272,
|
| 2333 |
+
"step": 3320
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 8.763157894736842,
|
| 2337 |
+
"grad_norm": 0.30955877900123596,
|
| 2338 |
+
"learning_rate": 4.124373819734795e-06,
|
| 2339 |
+
"loss": 0.0308,
|
| 2340 |
+
"step": 3330
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 8.789473684210526,
|
| 2344 |
+
"grad_norm": 0.210093155503273,
|
| 2345 |
+
"learning_rate": 3.953061475767339e-06,
|
| 2346 |
+
"loss": 0.0264,
|
| 2347 |
+
"step": 3340
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 8.81578947368421,
|
| 2351 |
+
"grad_norm": 0.29672837257385254,
|
| 2352 |
+
"learning_rate": 3.785236384038232e-06,
|
| 2353 |
+
"loss": 0.0295,
|
| 2354 |
+
"step": 3350
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 8.842105263157894,
|
| 2358 |
+
"grad_norm": 0.5261228680610657,
|
| 2359 |
+
"learning_rate": 3.620911254370224e-06,
|
| 2360 |
+
"loss": 0.0249,
|
| 2361 |
+
"step": 3360
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 8.868421052631579,
|
| 2365 |
+
"grad_norm": 0.22544153034687042,
|
| 2366 |
+
"learning_rate": 3.460098531525019e-06,
|
| 2367 |
+
"loss": 0.028,
|
| 2368 |
+
"step": 3370
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 8.894736842105264,
|
| 2372 |
+
"grad_norm": 0.24784667789936066,
|
| 2373 |
+
"learning_rate": 3.302810394260736e-06,
|
| 2374 |
+
"loss": 0.0252,
|
| 2375 |
+
"step": 3380
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 8.921052631578947,
|
| 2379 |
+
"grad_norm": 0.21960538625717163,
|
| 2380 |
+
"learning_rate": 3.1490587544096782e-06,
|
| 2381 |
+
"loss": 0.0255,
|
| 2382 |
+
"step": 3390
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 8.947368421052632,
|
| 2386 |
+
"grad_norm": 0.4938012361526489,
|
| 2387 |
+
"learning_rate": 2.9988552559761294e-06,
|
| 2388 |
+
"loss": 0.0261,
|
| 2389 |
+
"step": 3400
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 8.973684210526315,
|
| 2393 |
+
"grad_norm": 0.2430790662765503,
|
| 2394 |
+
"learning_rate": 2.85221127425459e-06,
|
| 2395 |
+
"loss": 0.0283,
|
| 2396 |
+
"step": 3410
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 9.0,
|
| 2400 |
+
"grad_norm": 0.3423117995262146,
|
| 2401 |
+
"learning_rate": 2.7091379149682685e-06,
|
| 2402 |
+
"loss": 0.0262,
|
| 2403 |
+
"step": 3420
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 9.026315789473685,
|
| 2407 |
+
"grad_norm": 0.1678830087184906,
|
| 2408 |
+
"learning_rate": 2.5696460134279955e-06,
|
| 2409 |
+
"loss": 0.0226,
|
| 2410 |
+
"step": 3430
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 9.052631578947368,
|
| 2414 |
+
"grad_norm": 0.1864800751209259,
|
| 2415 |
+
"learning_rate": 2.4337461337116894e-06,
|
| 2416 |
+
"loss": 0.028,
|
| 2417 |
+
"step": 3440
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 9.078947368421053,
|
| 2421 |
+
"grad_norm": 0.3498280942440033,
|
| 2422 |
+
"learning_rate": 2.3014485678642563e-06,
|
| 2423 |
+
"loss": 0.0251,
|
| 2424 |
+
"step": 3450
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 9.105263157894736,
|
| 2428 |
+
"grad_norm": 0.2959420680999756,
|
| 2429 |
+
"learning_rate": 2.1727633351182e-06,
|
| 2430 |
+
"loss": 0.0267,
|
| 2431 |
+
"step": 3460
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 9.131578947368421,
|
| 2435 |
+
"grad_norm": 0.46708524227142334,
|
| 2436 |
+
"learning_rate": 2.0477001811347985e-06,
|
| 2437 |
+
"loss": 0.0271,
|
| 2438 |
+
"step": 3470
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 9.157894736842104,
|
| 2442 |
+
"grad_norm": 0.21950027346611023,
|
| 2443 |
+
"learning_rate": 1.9262685772660606e-06,
|
| 2444 |
+
"loss": 0.0241,
|
| 2445 |
+
"step": 3480
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 9.18421052631579,
|
| 2449 |
+
"grad_norm": 0.19924266636371613,
|
| 2450 |
+
"learning_rate": 1.8084777198374315e-06,
|
| 2451 |
+
"loss": 0.0276,
|
| 2452 |
+
"step": 3490
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 9.210526315789474,
|
| 2456 |
+
"grad_norm": 0.29325613379478455,
|
| 2457 |
+
"learning_rate": 1.6943365294513236e-06,
|
| 2458 |
+
"loss": 0.0278,
|
| 2459 |
+
"step": 3500
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 9.236842105263158,
|
| 2463 |
+
"grad_norm": 0.19263319671154022,
|
| 2464 |
+
"learning_rate": 1.5838536503115675e-06,
|
| 2465 |
+
"loss": 0.0275,
|
| 2466 |
+
"step": 3510
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 9.263157894736842,
|
| 2470 |
+
"grad_norm": 0.3293076753616333,
|
| 2471 |
+
"learning_rate": 1.4770374495687134e-06,
|
| 2472 |
+
"loss": 0.0232,
|
| 2473 |
+
"step": 3520
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 9.289473684210526,
|
| 2477 |
+
"grad_norm": 0.23276633024215698,
|
| 2478 |
+
"learning_rate": 1.3738960166864101e-06,
|
| 2479 |
+
"loss": 0.0245,
|
| 2480 |
+
"step": 3530
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 9.31578947368421,
|
| 2484 |
+
"grad_norm": 0.3246747553348541,
|
| 2485 |
+
"learning_rate": 1.274437162828751e-06,
|
| 2486 |
+
"loss": 0.0258,
|
| 2487 |
+
"step": 3540
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 9.342105263157896,
|
| 2491 |
+
"grad_norm": 0.3524140417575836,
|
| 2492 |
+
"learning_rate": 1.1786684202687026e-06,
|
| 2493 |
+
"loss": 0.023,
|
| 2494 |
+
"step": 3550
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 9.368421052631579,
|
| 2498 |
+
"grad_norm": 0.27717113494873047,
|
| 2499 |
+
"learning_rate": 1.0865970418177051e-06,
|
| 2500 |
+
"loss": 0.0292,
|
| 2501 |
+
"step": 3560
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 9.394736842105264,
|
| 2505 |
+
"grad_norm": 0.29303857684135437,
|
| 2506 |
+
"learning_rate": 9.98230000276351e-07,
|
| 2507 |
+
"loss": 0.0239,
|
| 2508 |
+
"step": 3570
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 9.421052631578947,
|
| 2512 |
+
"grad_norm": 0.5621191263198853,
|
| 2513 |
+
"learning_rate": 9.135739879063465e-07,
|
| 2514 |
+
"loss": 0.0242,
|
| 2515 |
+
"step": 3580
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 9.447368421052632,
|
| 2519 |
+
"grad_norm": 0.41081079840660095,
|
| 2520 |
+
"learning_rate": 8.326354159236882e-07,
|
| 2521 |
+
"loss": 0.0232,
|
| 2522 |
+
"step": 3590
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 9.473684210526315,
|
| 2526 |
+
"grad_norm": 0.48332345485687256,
|
| 2527 |
+
"learning_rate": 7.554204140131138e-07,
|
| 2528 |
+
"loss": 0.026,
|
| 2529 |
+
"step": 3600
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 9.5,
|
| 2533 |
+
"grad_norm": 0.36336687207221985,
|
| 2534 |
+
"learning_rate": 6.819348298638839e-07,
|
| 2535 |
+
"loss": 0.0226,
|
| 2536 |
+
"step": 3610
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 9.526315789473685,
|
| 2540 |
+
"grad_norm": 0.4939121901988983,
|
| 2541 |
+
"learning_rate": 6.121842287269419e-07,
|
| 2542 |
+
"loss": 0.0274,
|
| 2543 |
+
"step": 3620
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 9.552631578947368,
|
| 2547 |
+
"grad_norm": 0.60195392370224,
|
| 2548 |
+
"learning_rate": 5.46173892993429e-07,
|
| 2549 |
+
"loss": 0.0229,
|
| 2550 |
+
"step": 3630
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 9.578947368421053,
|
| 2554 |
+
"grad_norm": 0.3372774124145508,
|
| 2555 |
+
"learning_rate": 4.839088217946208e-07,
|
| 2556 |
+
"loss": 0.0236,
|
| 2557 |
+
"step": 3640
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 9.605263157894736,
|
| 2561 |
+
"grad_norm": 0.2751666307449341,
|
| 2562 |
+
"learning_rate": 4.253937306233691e-07,
|
| 2563 |
+
"loss": 0.0267,
|
| 2564 |
+
"step": 3650
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 9.631578947368421,
|
| 2568 |
+
"grad_norm": 0.2811025381088257,
|
| 2569 |
+
"learning_rate": 3.706330509769429e-07,
|
| 2570 |
+
"loss": 0.0269,
|
| 2571 |
+
"step": 3660
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 9.657894736842106,
|
| 2575 |
+
"grad_norm": 0.27488309144973755,
|
| 2576 |
+
"learning_rate": 3.1963093002145285e-07,
|
| 2577 |
+
"loss": 0.0294,
|
| 2578 |
+
"step": 3670
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 9.68421052631579,
|
| 2582 |
+
"grad_norm": 0.29794642329216003,
|
| 2583 |
+
"learning_rate": 2.7239123027775204e-07,
|
| 2584 |
+
"loss": 0.0226,
|
| 2585 |
+
"step": 3680
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 9.710526315789474,
|
| 2589 |
+
"grad_norm": 0.49477118253707886,
|
| 2590 |
+
"learning_rate": 2.289175293289314e-07,
|
| 2591 |
+
"loss": 0.0263,
|
| 2592 |
+
"step": 3690
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 9.736842105263158,
|
| 2596 |
+
"grad_norm": 0.44544729590415955,
|
| 2597 |
+
"learning_rate": 1.8921311954937516e-07,
|
| 2598 |
+
"loss": 0.024,
|
| 2599 |
+
"step": 3700
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 9.763157894736842,
|
| 2603 |
+
"grad_norm": 0.380344033241272,
|
| 2604 |
+
"learning_rate": 1.5328100785542697e-07,
|
| 2605 |
+
"loss": 0.0243,
|
| 2606 |
+
"step": 3710
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 9.789473684210526,
|
| 2610 |
+
"grad_norm": 0.2532555162906647,
|
| 2611 |
+
"learning_rate": 1.211239154776611e-07,
|
| 2612 |
+
"loss": 0.0256,
|
| 2613 |
+
"step": 3720
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 9.81578947368421,
|
| 2617 |
+
"grad_norm": 0.3260039985179901,
|
| 2618 |
+
"learning_rate": 9.27442777547971e-08,
|
| 2619 |
+
"loss": 0.0253,
|
| 2620 |
+
"step": 3730
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 9.842105263157894,
|
| 2624 |
+
"grad_norm": 0.2649020254611969,
|
| 2625 |
+
"learning_rate": 6.814424394926966e-08,
|
| 2626 |
+
"loss": 0.0298,
|
| 2627 |
+
"step": 3740
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 9.868421052631579,
|
| 2631 |
+
"grad_norm": 0.257893443107605,
|
| 2632 |
+
"learning_rate": 4.732567708445878e-08,
|
| 2633 |
+
"loss": 0.0223,
|
| 2634 |
+
"step": 3750
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 9.894736842105264,
|
| 2638 |
+
"grad_norm": 0.20920208096504211,
|
| 2639 |
+
"learning_rate": 3.029015380359157e-08,
|
| 2640 |
+
"loss": 0.0249,
|
| 2641 |
+
"step": 3760
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 9.921052631578947,
|
| 2645 |
+
"grad_norm": 0.19053097069263458,
|
| 2646 |
+
"learning_rate": 1.7038964250343238e-08,
|
| 2647 |
+
"loss": 0.029,
|
| 2648 |
+
"step": 3770
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 9.947368421052632,
|
| 2652 |
+
"grad_norm": 0.3256742060184479,
|
| 2653 |
+
"learning_rate": 7.573111971148627e-09,
|
| 2654 |
+
"loss": 0.0254,
|
| 2655 |
+
"step": 3780
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 9.973684210526315,
|
| 2659 |
+
"grad_norm": 0.3680499792098999,
|
| 2660 |
+
"learning_rate": 1.8933138391574732e-09,
|
| 2661 |
+
"loss": 0.025,
|
| 2662 |
+
"step": 3790
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 10.0,
|
| 2666 |
+
"grad_norm": 0.5628379583358765,
|
| 2667 |
+
"learning_rate": 0.0,
|
| 2668 |
+
"loss": 0.0263,
|
| 2669 |
+
"step": 3800
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 10.0,
|
| 2673 |
+
"step": 3800,
|
| 2674 |
+
"total_flos": 4.0064924863782144e+17,
|
| 2675 |
+
"train_loss": 0.05822706136264299,
|
| 2676 |
+
"train_runtime": 3981.0494,
|
| 2677 |
+
"train_samples_per_second": 46.716,
|
| 2678 |
+
"train_steps_per_second": 0.955
|
| 2679 |
+
}
|
| 2680 |
+
],
|
| 2681 |
+
"logging_steps": 10,
|
| 2682 |
+
"max_steps": 3800,
|
| 2683 |
+
"num_input_tokens_seen": 0,
|
| 2684 |
+
"num_train_epochs": 10,
|
| 2685 |
+
"save_steps": 10000,
|
| 2686 |
+
"stateful_callbacks": {
|
| 2687 |
+
"TrainerControl": {
|
| 2688 |
+
"args": {
|
| 2689 |
+
"should_epoch_stop": false,
|
| 2690 |
+
"should_evaluate": false,
|
| 2691 |
+
"should_log": false,
|
| 2692 |
+
"should_save": true,
|
| 2693 |
+
"should_training_stop": true
|
| 2694 |
+
},
|
| 2695 |
+
"attributes": {}
|
| 2696 |
+
}
|
| 2697 |
+
},
|
| 2698 |
+
"total_flos": 4.0064924863782144e+17,
|
| 2699 |
+
"train_batch_size": 49,
|
| 2700 |
+
"trial_name": null,
|
| 2701 |
+
"trial_params": null
|
| 2702 |
+
}
|