Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2414 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2414 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.2956521739130435,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3400,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0008695652173913044,
|
| 14 |
+
"grad_norm": 0.0412801168859005,
|
| 15 |
+
"learning_rate": 7.875e-06,
|
| 16 |
+
"loss": 1.986367416381836,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0017391304347826088,
|
| 21 |
+
"grad_norm": 0.04147534817457199,
|
| 22 |
+
"learning_rate": 1.6625e-05,
|
| 23 |
+
"loss": 1.8889337539672852,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0026086956521739132,
|
| 28 |
+
"grad_norm": 0.041172582656145096,
|
| 29 |
+
"learning_rate": 2.5374999999999997e-05,
|
| 30 |
+
"loss": 1.9236927032470703,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.0034782608695652175,
|
| 35 |
+
"grad_norm": 0.041271984577178955,
|
| 36 |
+
"learning_rate": 3.4125e-05,
|
| 37 |
+
"loss": 1.836932373046875,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.004347826086956522,
|
| 42 |
+
"grad_norm": 0.04144110530614853,
|
| 43 |
+
"learning_rate": 4.2875e-05,
|
| 44 |
+
"loss": 1.8578208923339843,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.0052173913043478265,
|
| 49 |
+
"grad_norm": 0.04153507575392723,
|
| 50 |
+
"learning_rate": 5.1624999999999994e-05,
|
| 51 |
+
"loss": 1.9032476425170899,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.00608695652173913,
|
| 56 |
+
"grad_norm": 0.04170163720846176,
|
| 57 |
+
"learning_rate": 6.0375e-05,
|
| 58 |
+
"loss": 1.956137466430664,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.006956521739130435,
|
| 63 |
+
"grad_norm": 0.04149318486452103,
|
| 64 |
+
"learning_rate": 6.9125e-05,
|
| 65 |
+
"loss": 1.9041580200195312,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.00782608695652174,
|
| 70 |
+
"grad_norm": 0.04267331585288048,
|
| 71 |
+
"learning_rate": 7.7875e-05,
|
| 72 |
+
"loss": 1.8978221893310547,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.008695652173913044,
|
| 77 |
+
"grad_norm": 0.04074378311634064,
|
| 78 |
+
"learning_rate": 8.6625e-05,
|
| 79 |
+
"loss": 1.8678691864013672,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.009565217391304347,
|
| 84 |
+
"grad_norm": 0.0422818660736084,
|
| 85 |
+
"learning_rate": 9.5375e-05,
|
| 86 |
+
"loss": 1.868328857421875,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.010434782608695653,
|
| 91 |
+
"grad_norm": 0.041968777775764465,
|
| 92 |
+
"learning_rate": 0.000104125,
|
| 93 |
+
"loss": 1.891305923461914,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.011304347826086957,
|
| 98 |
+
"grad_norm": 0.03967316821217537,
|
| 99 |
+
"learning_rate": 0.000112875,
|
| 100 |
+
"loss": 1.8953823089599608,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.01217391304347826,
|
| 105 |
+
"grad_norm": 0.04263008385896683,
|
| 106 |
+
"learning_rate": 0.00012162499999999999,
|
| 107 |
+
"loss": 1.8702178955078126,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.013043478260869565,
|
| 112 |
+
"grad_norm": 0.042412299662828445,
|
| 113 |
+
"learning_rate": 0.000130375,
|
| 114 |
+
"loss": 1.8251636505126954,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.01391304347826087,
|
| 119 |
+
"grad_norm": 0.04295488819479942,
|
| 120 |
+
"learning_rate": 0.000139125,
|
| 121 |
+
"loss": 1.9118040084838868,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.014782608695652174,
|
| 126 |
+
"grad_norm": 0.043725475668907166,
|
| 127 |
+
"learning_rate": 0.000147875,
|
| 128 |
+
"loss": 1.939938735961914,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.01565217391304348,
|
| 133 |
+
"grad_norm": 0.04236534982919693,
|
| 134 |
+
"learning_rate": 0.000156625,
|
| 135 |
+
"loss": 1.9232837677001953,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.01652173913043478,
|
| 140 |
+
"grad_norm": 0.04496081545948982,
|
| 141 |
+
"learning_rate": 0.000165375,
|
| 142 |
+
"loss": 1.979838752746582,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.017391304347826087,
|
| 147 |
+
"grad_norm": 0.043251801282167435,
|
| 148 |
+
"learning_rate": 0.000174125,
|
| 149 |
+
"loss": 1.9420089721679688,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.018260869565217393,
|
| 154 |
+
"grad_norm": 0.04351160675287247,
|
| 155 |
+
"learning_rate": 0.0001749997260913748,
|
| 156 |
+
"loss": 1.8539535522460937,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.019130434782608695,
|
| 161 |
+
"grad_norm": 0.04367993399500847,
|
| 162 |
+
"learning_rate": 0.00017499877924894623,
|
| 163 |
+
"loss": 1.8686515808105468,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.02,
|
| 168 |
+
"grad_norm": 0.04434121400117874,
|
| 169 |
+
"learning_rate": 0.000174997156098443,
|
| 170 |
+
"loss": 1.8759822845458984,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.020869565217391306,
|
| 175 |
+
"grad_norm": 0.04514431208372116,
|
| 176 |
+
"learning_rate": 0.000174994856652411,
|
| 177 |
+
"loss": 1.9328535079956055,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.021739130434782608,
|
| 182 |
+
"grad_norm": 0.04443835839629173,
|
| 183 |
+
"learning_rate": 0.00017499188092862342,
|
| 184 |
+
"loss": 1.9367387771606446,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.022608695652173914,
|
| 189 |
+
"grad_norm": 0.04333459958434105,
|
| 190 |
+
"learning_rate": 0.00017498822895008068,
|
| 191 |
+
"loss": 1.9187299728393554,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.023478260869565216,
|
| 196 |
+
"grad_norm": 0.04284181818366051,
|
| 197 |
+
"learning_rate": 0.00017498390074501016,
|
| 198 |
+
"loss": 1.8823593139648438,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.02434782608695652,
|
| 203 |
+
"grad_norm": 0.042989395558834076,
|
| 204 |
+
"learning_rate": 0.00017497889634686607,
|
| 205 |
+
"loss": 1.9293874740600585,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.025217391304347827,
|
| 210 |
+
"grad_norm": 0.04543396085500717,
|
| 211 |
+
"learning_rate": 0.00017497321579432908,
|
| 212 |
+
"loss": 1.9312299728393554,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.02608695652173913,
|
| 217 |
+
"grad_norm": 0.045147597789764404,
|
| 218 |
+
"learning_rate": 0.00017496685913130617,
|
| 219 |
+
"loss": 1.905094528198242,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.026956521739130435,
|
| 224 |
+
"grad_norm": 0.04503512755036354,
|
| 225 |
+
"learning_rate": 0.0001749598264069301,
|
| 226 |
+
"loss": 1.9508794784545898,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.02782608695652174,
|
| 231 |
+
"grad_norm": 0.04418083652853966,
|
| 232 |
+
"learning_rate": 0.00017495211767555927,
|
| 233 |
+
"loss": 1.93798828125,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.028695652173913042,
|
| 238 |
+
"grad_norm": 0.0446457639336586,
|
| 239 |
+
"learning_rate": 0.0001749437329967771,
|
| 240 |
+
"loss": 1.8692955017089843,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.029565217391304348,
|
| 245 |
+
"grad_norm": 0.045488957315683365,
|
| 246 |
+
"learning_rate": 0.00017493467243539165,
|
| 247 |
+
"loss": 1.927821731567383,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.030434782608695653,
|
| 252 |
+
"grad_norm": 0.04600228741765022,
|
| 253 |
+
"learning_rate": 0.00017492493606143506,
|
| 254 |
+
"loss": 1.9788978576660157,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.03130434782608696,
|
| 259 |
+
"grad_norm": 0.044314391911029816,
|
| 260 |
+
"learning_rate": 0.00017491452395016318,
|
| 261 |
+
"loss": 1.8822463989257812,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.03217391304347826,
|
| 266 |
+
"grad_norm": 0.04477265104651451,
|
| 267 |
+
"learning_rate": 0.00017490343618205474,
|
| 268 |
+
"loss": 1.923102569580078,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.03304347826086956,
|
| 273 |
+
"grad_norm": 0.04579043760895729,
|
| 274 |
+
"learning_rate": 0.0001748916728428109,
|
| 275 |
+
"loss": 1.9119430541992188,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.03391304347826087,
|
| 280 |
+
"grad_norm": 0.04229314625263214,
|
| 281 |
+
"learning_rate": 0.00017487923402335453,
|
| 282 |
+
"loss": 1.9644861221313477,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.034782608695652174,
|
| 287 |
+
"grad_norm": 0.04624266177415848,
|
| 288 |
+
"learning_rate": 0.00017486611981982962,
|
| 289 |
+
"loss": 1.934438705444336,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.035652173913043476,
|
| 294 |
+
"grad_norm": 0.044226158410310745,
|
| 295 |
+
"learning_rate": 0.00017485233033360026,
|
| 296 |
+
"loss": 1.9088932037353517,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.036521739130434785,
|
| 301 |
+
"grad_norm": 0.0430554524064064,
|
| 302 |
+
"learning_rate": 0.0001748378656712501,
|
| 303 |
+
"loss": 1.9586982727050781,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.03739130434782609,
|
| 308 |
+
"grad_norm": 0.04383225739002228,
|
| 309 |
+
"learning_rate": 0.00017482272594458156,
|
| 310 |
+
"loss": 1.9529533386230469,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.03826086956521739,
|
| 315 |
+
"grad_norm": 0.04541724920272827,
|
| 316 |
+
"learning_rate": 0.00017480691127061475,
|
| 317 |
+
"loss": 1.9128194808959962,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.0391304347826087,
|
| 322 |
+
"grad_norm": 0.045937202870845795,
|
| 323 |
+
"learning_rate": 0.00017479042177158665,
|
| 324 |
+
"loss": 1.9061151504516602,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.04,
|
| 329 |
+
"grad_norm": 0.04901459440588951,
|
| 330 |
+
"learning_rate": 0.00017477325757495036,
|
| 331 |
+
"loss": 1.9450111389160156,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.0408695652173913,
|
| 336 |
+
"grad_norm": 0.04544888064265251,
|
| 337 |
+
"learning_rate": 0.00017475541881337373,
|
| 338 |
+
"loss": 1.8662300109863281,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.04173913043478261,
|
| 343 |
+
"grad_norm": 0.044371675699949265,
|
| 344 |
+
"learning_rate": 0.00017473690562473874,
|
| 345 |
+
"loss": 1.9551380157470704,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.042608695652173914,
|
| 350 |
+
"grad_norm": 0.04671357944607735,
|
| 351 |
+
"learning_rate": 0.00017471771815214017,
|
| 352 |
+
"loss": 1.941476821899414,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.043478260869565216,
|
| 357 |
+
"grad_norm": 0.04723666235804558,
|
| 358 |
+
"learning_rate": 0.00017469785654388454,
|
| 359 |
+
"loss": 1.9640892028808594,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.04434782608695652,
|
| 364 |
+
"grad_norm": 0.04654547944664955,
|
| 365 |
+
"learning_rate": 0.00017467732095348905,
|
| 366 |
+
"loss": 1.944451141357422,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.04521739130434783,
|
| 371 |
+
"grad_norm": 0.0464133694767952,
|
| 372 |
+
"learning_rate": 0.00017465611153968035,
|
| 373 |
+
"loss": 1.9126354217529298,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.04608695652173913,
|
| 378 |
+
"grad_norm": 0.04279818385839462,
|
| 379 |
+
"learning_rate": 0.00017463422846639325,
|
| 380 |
+
"loss": 1.9004449844360352,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.04695652173913043,
|
| 385 |
+
"grad_norm": 0.045143257826566696,
|
| 386 |
+
"learning_rate": 0.0001746116719027695,
|
| 387 |
+
"loss": 1.8843421936035156,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.04782608695652174,
|
| 392 |
+
"grad_norm": 0.04402666911482811,
|
| 393 |
+
"learning_rate": 0.00017458844202315656,
|
| 394 |
+
"loss": 1.9196475982666015,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.04869565217391304,
|
| 399 |
+
"grad_norm": 0.04636216163635254,
|
| 400 |
+
"learning_rate": 0.0001745645390071061,
|
| 401 |
+
"loss": 1.9364383697509766,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.049565217391304345,
|
| 406 |
+
"grad_norm": 0.044472336769104004,
|
| 407 |
+
"learning_rate": 0.00017453996303937278,
|
| 408 |
+
"loss": 1.9093303680419922,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.050434782608695654,
|
| 413 |
+
"grad_norm": 0.04402562230825424,
|
| 414 |
+
"learning_rate": 0.00017451471430991261,
|
| 415 |
+
"loss": 1.9382781982421875,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.051304347826086956,
|
| 420 |
+
"grad_norm": 0.045037273317575455,
|
| 421 |
+
"learning_rate": 0.0001744887930138818,
|
| 422 |
+
"loss": 1.9814619064331054,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.05217391304347826,
|
| 427 |
+
"grad_norm": 0.04517680034041405,
|
| 428 |
+
"learning_rate": 0.0001744621993516348,
|
| 429 |
+
"loss": 1.8577213287353516,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.05304347826086957,
|
| 434 |
+
"grad_norm": 0.044585905969142914,
|
| 435 |
+
"learning_rate": 0.0001744349335287232,
|
| 436 |
+
"loss": 1.906428909301758,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.05391304347826087,
|
| 441 |
+
"grad_norm": 0.047429412603378296,
|
| 442 |
+
"learning_rate": 0.00017440699575589392,
|
| 443 |
+
"loss": 1.956072998046875,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.05478260869565217,
|
| 448 |
+
"grad_norm": 0.04587710648775101,
|
| 449 |
+
"learning_rate": 0.0001743783862490876,
|
| 450 |
+
"loss": 1.9825408935546875,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.05565217391304348,
|
| 455 |
+
"grad_norm": 0.044858191162347794,
|
| 456 |
+
"learning_rate": 0.00017434910522943687,
|
| 457 |
+
"loss": 1.9249252319335937,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.05652173913043478,
|
| 462 |
+
"grad_norm": 0.046517107635736465,
|
| 463 |
+
"learning_rate": 0.0001743191529232648,
|
| 464 |
+
"loss": 1.9157310485839845,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.057391304347826085,
|
| 469 |
+
"grad_norm": 0.04853269085288048,
|
| 470 |
+
"learning_rate": 0.00017428852956208308,
|
| 471 |
+
"loss": 1.9744232177734375,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.058260869565217394,
|
| 476 |
+
"grad_norm": 0.04673819616436958,
|
| 477 |
+
"learning_rate": 0.0001742572353825901,
|
| 478 |
+
"loss": 1.984130859375,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.059130434782608696,
|
| 483 |
+
"grad_norm": 0.0440078005194664,
|
| 484 |
+
"learning_rate": 0.00017422527062666934,
|
| 485 |
+
"loss": 1.841623306274414,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.06,
|
| 490 |
+
"grad_norm": 0.04755179211497307,
|
| 491 |
+
"learning_rate": 0.00017419263554138738,
|
| 492 |
+
"loss": 1.8390884399414062,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.06086956521739131,
|
| 497 |
+
"grad_norm": 0.04695327207446098,
|
| 498 |
+
"learning_rate": 0.00017415933037899197,
|
| 499 |
+
"loss": 1.9590534210205077,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.06173913043478261,
|
| 504 |
+
"grad_norm": 0.04554755613207817,
|
| 505 |
+
"learning_rate": 0.00017412535539691013,
|
| 506 |
+
"loss": 1.9262825012207032,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.06260869565217392,
|
| 511 |
+
"grad_norm": 0.04517935588955879,
|
| 512 |
+
"learning_rate": 0.00017409071085774613,
|
| 513 |
+
"loss": 1.9321512222290038,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.06347826086956522,
|
| 518 |
+
"grad_norm": 0.045931559056043625,
|
| 519 |
+
"learning_rate": 0.00017405539702927954,
|
| 520 |
+
"loss": 1.8974014282226563,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.06434782608695652,
|
| 525 |
+
"grad_norm": 0.04443606361746788,
|
| 526 |
+
"learning_rate": 0.00017401941418446301,
|
| 527 |
+
"loss": 1.9469221115112305,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.06521739130434782,
|
| 532 |
+
"grad_norm": 0.04661998152732849,
|
| 533 |
+
"learning_rate": 0.00017398276260142034,
|
| 534 |
+
"loss": 1.9025968551635741,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.06608695652173913,
|
| 539 |
+
"grad_norm": 0.044536832720041275,
|
| 540 |
+
"learning_rate": 0.00017394544256344412,
|
| 541 |
+
"loss": 1.9783390045166016,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.06695652173913043,
|
| 546 |
+
"grad_norm": 0.04502754285931587,
|
| 547 |
+
"learning_rate": 0.0001739074543589938,
|
| 548 |
+
"loss": 1.9079425811767579,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.06782608695652174,
|
| 553 |
+
"grad_norm": 0.045174986124038696,
|
| 554 |
+
"learning_rate": 0.00017386879828169315,
|
| 555 |
+
"loss": 1.9875041961669921,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.06869565217391305,
|
| 560 |
+
"grad_norm": 0.04711156338453293,
|
| 561 |
+
"learning_rate": 0.00017382947463032833,
|
| 562 |
+
"loss": 1.9256580352783204,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.06956521739130435,
|
| 567 |
+
"grad_norm": 0.046600691974163055,
|
| 568 |
+
"learning_rate": 0.00017378948370884526,
|
| 569 |
+
"loss": 1.9062789916992187,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.07043478260869565,
|
| 574 |
+
"grad_norm": 0.04497547820210457,
|
| 575 |
+
"learning_rate": 0.00017374882582634756,
|
| 576 |
+
"loss": 1.9336435317993164,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.07130434782608695,
|
| 581 |
+
"grad_norm": 0.045298535376787186,
|
| 582 |
+
"learning_rate": 0.00017370750129709392,
|
| 583 |
+
"loss": 1.9081130981445313,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.07217391304347825,
|
| 588 |
+
"grad_norm": 0.04361579939723015,
|
| 589 |
+
"learning_rate": 0.00017366551044049576,
|
| 590 |
+
"loss": 1.9321014404296875,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.07304347826086957,
|
| 595 |
+
"grad_norm": 0.04433725029230118,
|
| 596 |
+
"learning_rate": 0.00017362285358111485,
|
| 597 |
+
"loss": 1.8640398025512694,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.07391304347826087,
|
| 602 |
+
"grad_norm": 0.04579870030283928,
|
| 603 |
+
"learning_rate": 0.0001735795310486607,
|
| 604 |
+
"loss": 1.9060062408447265,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.07478260869565218,
|
| 609 |
+
"grad_norm": 0.04460195451974869,
|
| 610 |
+
"learning_rate": 0.000173535543177988,
|
| 611 |
+
"loss": 1.9636484146118165,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.07565217391304348,
|
| 616 |
+
"grad_norm": 0.046220432966947556,
|
| 617 |
+
"learning_rate": 0.00017349089030909407,
|
| 618 |
+
"loss": 1.928147315979004,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.07652173913043478,
|
| 623 |
+
"grad_norm": 0.04606110602617264,
|
| 624 |
+
"learning_rate": 0.00017344557278711628,
|
| 625 |
+
"loss": 1.9536296844482421,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.07739130434782608,
|
| 630 |
+
"grad_norm": 0.045817017555236816,
|
| 631 |
+
"learning_rate": 0.00017339959096232928,
|
| 632 |
+
"loss": 1.9305814743041991,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.0782608695652174,
|
| 637 |
+
"grad_norm": 0.04592573270201683,
|
| 638 |
+
"learning_rate": 0.00017335294519014243,
|
| 639 |
+
"loss": 1.8813438415527344,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.0791304347826087,
|
| 644 |
+
"grad_norm": 0.04669544845819473,
|
| 645 |
+
"learning_rate": 0.0001733056358310968,
|
| 646 |
+
"loss": 1.9484172821044923,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.08,
|
| 651 |
+
"grad_norm": 0.04449376091361046,
|
| 652 |
+
"learning_rate": 0.0001732576632508627,
|
| 653 |
+
"loss": 1.8734165191650392,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.0808695652173913,
|
| 658 |
+
"grad_norm": 0.04569607973098755,
|
| 659 |
+
"learning_rate": 0.00017320902782023666,
|
| 660 |
+
"loss": 1.904270553588867,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.0817391304347826,
|
| 665 |
+
"grad_norm": 0.047245677560567856,
|
| 666 |
+
"learning_rate": 0.00017315972991513857,
|
| 667 |
+
"loss": 1.944266128540039,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.08260869565217391,
|
| 672 |
+
"grad_norm": 0.04679643735289574,
|
| 673 |
+
"learning_rate": 0.00017310976991660891,
|
| 674 |
+
"loss": 1.9376203536987304,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.08347826086956522,
|
| 679 |
+
"grad_norm": 0.04880222678184509,
|
| 680 |
+
"learning_rate": 0.00017305914821080562,
|
| 681 |
+
"loss": 2.0033763885498046,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.08434782608695653,
|
| 686 |
+
"grad_norm": 0.04653260484337807,
|
| 687 |
+
"learning_rate": 0.0001730078651890012,
|
| 688 |
+
"loss": 1.877172088623047,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.08521739130434783,
|
| 693 |
+
"grad_norm": 0.04537530615925789,
|
| 694 |
+
"learning_rate": 0.00017295592124757974,
|
| 695 |
+
"loss": 1.8832229614257812,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.08608695652173913,
|
| 700 |
+
"grad_norm": 0.0438741110265255,
|
| 701 |
+
"learning_rate": 0.00017290331678803373,
|
| 702 |
+
"loss": 1.8948379516601563,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.08695652173913043,
|
| 707 |
+
"grad_norm": 0.04543806612491608,
|
| 708 |
+
"learning_rate": 0.00017285005221696114,
|
| 709 |
+
"loss": 1.8686225891113282,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.08782608695652173,
|
| 714 |
+
"grad_norm": 0.044838301837444305,
|
| 715 |
+
"learning_rate": 0.0001727961279460621,
|
| 716 |
+
"loss": 1.9170583724975585,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.08869565217391304,
|
| 721 |
+
"grad_norm": 0.04491033777594566,
|
| 722 |
+
"learning_rate": 0.00017274154439213577,
|
| 723 |
+
"loss": 1.9348651885986328,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.08956521739130435,
|
| 728 |
+
"grad_norm": 0.045102160423994064,
|
| 729 |
+
"learning_rate": 0.00017268630197707715,
|
| 730 |
+
"loss": 1.9238882064819336,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.09043478260869565,
|
| 735 |
+
"grad_norm": 0.044371772557497025,
|
| 736 |
+
"learning_rate": 0.00017263040112787387,
|
| 737 |
+
"loss": 1.8916519165039063,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.09130434782608696,
|
| 742 |
+
"grad_norm": 0.04538256302475929,
|
| 743 |
+
"learning_rate": 0.00017257384227660271,
|
| 744 |
+
"loss": 1.8470991134643555,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.09217391304347826,
|
| 749 |
+
"grad_norm": 0.046616047620773315,
|
| 750 |
+
"learning_rate": 0.00017251662586042644,
|
| 751 |
+
"loss": 1.9685466766357422,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.09304347826086956,
|
| 756 |
+
"grad_norm": 0.04723796248435974,
|
| 757 |
+
"learning_rate": 0.0001724587523215904,
|
| 758 |
+
"loss": 1.8623455047607422,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.09391304347826086,
|
| 763 |
+
"grad_norm": 0.04548574984073639,
|
| 764 |
+
"learning_rate": 0.00017240022210741893,
|
| 765 |
+
"loss": 1.8760478973388672,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.09478260869565218,
|
| 770 |
+
"grad_norm": 0.04325335472822189,
|
| 771 |
+
"learning_rate": 0.00017234103567031225,
|
| 772 |
+
"loss": 1.839696502685547,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.09565217391304348,
|
| 777 |
+
"grad_norm": 0.04418843984603882,
|
| 778 |
+
"learning_rate": 0.00017228119346774257,
|
| 779 |
+
"loss": 1.9864925384521483,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.09652173913043478,
|
| 784 |
+
"grad_norm": 0.04545409604907036,
|
| 785 |
+
"learning_rate": 0.00017222069596225078,
|
| 786 |
+
"loss": 1.9569078445434571,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.09739130434782609,
|
| 791 |
+
"grad_norm": 0.047114428132772446,
|
| 792 |
+
"learning_rate": 0.00017215954362144285,
|
| 793 |
+
"loss": 1.94586181640625,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.09826086956521739,
|
| 798 |
+
"grad_norm": 0.046218354254961014,
|
| 799 |
+
"learning_rate": 0.00017209773691798628,
|
| 800 |
+
"loss": 1.959282684326172,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.09913043478260869,
|
| 805 |
+
"grad_norm": 0.04634036496281624,
|
| 806 |
+
"learning_rate": 0.00017203527632960619,
|
| 807 |
+
"loss": 1.8650602340698241,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.1,
|
| 812 |
+
"grad_norm": 0.048119544982910156,
|
| 813 |
+
"learning_rate": 0.0001719721623390819,
|
| 814 |
+
"loss": 1.9122610092163086,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.10086956521739131,
|
| 819 |
+
"grad_norm": 0.045855700969696045,
|
| 820 |
+
"learning_rate": 0.00017190839543424312,
|
| 821 |
+
"loss": 1.9363866806030274,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.10173913043478261,
|
| 826 |
+
"grad_norm": 0.04780857637524605,
|
| 827 |
+
"learning_rate": 0.0001718439761079661,
|
| 828 |
+
"loss": 1.9255781173706055,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.10260869565217391,
|
| 833 |
+
"grad_norm": 0.044750381261110306,
|
| 834 |
+
"learning_rate": 0.00017177890485816993,
|
| 835 |
+
"loss": 1.8942161560058595,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.10347826086956521,
|
| 840 |
+
"grad_norm": 0.0476827546954155,
|
| 841 |
+
"learning_rate": 0.0001717131821878126,
|
| 842 |
+
"loss": 1.9411766052246093,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.10434782608695652,
|
| 847 |
+
"grad_norm": 0.046074964106082916,
|
| 848 |
+
"learning_rate": 0.0001716468086048872,
|
| 849 |
+
"loss": 1.8827497482299804,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.10521739130434783,
|
| 854 |
+
"grad_norm": 0.04519912600517273,
|
| 855 |
+
"learning_rate": 0.0001715797846224179,
|
| 856 |
+
"loss": 1.9211563110351562,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.10608695652173913,
|
| 861 |
+
"grad_norm": 0.04722365736961365,
|
| 862 |
+
"learning_rate": 0.00017151211075845603,
|
| 863 |
+
"loss": 1.9217552185058593,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.10695652173913044,
|
| 868 |
+
"grad_norm": 0.046000994741916656,
|
| 869 |
+
"learning_rate": 0.00017144378753607608,
|
| 870 |
+
"loss": 1.968286895751953,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.10782608695652174,
|
| 875 |
+
"grad_norm": 0.04773630201816559,
|
| 876 |
+
"learning_rate": 0.00017137481548337174,
|
| 877 |
+
"loss": 1.9427530288696289,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.10869565217391304,
|
| 882 |
+
"grad_norm": 0.04625891149044037,
|
| 883 |
+
"learning_rate": 0.0001713051951334516,
|
| 884 |
+
"loss": 1.8612529754638671,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.10956521739130434,
|
| 889 |
+
"grad_norm": 0.04403012990951538,
|
| 890 |
+
"learning_rate": 0.00017123492702443522,
|
| 891 |
+
"loss": 1.91204776763916,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.11043478260869566,
|
| 896 |
+
"grad_norm": 0.04643299803137779,
|
| 897 |
+
"learning_rate": 0.00017116401169944896,
|
| 898 |
+
"loss": 1.9106037139892578,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.11130434782608696,
|
| 903 |
+
"grad_norm": 0.04438823461532593,
|
| 904 |
+
"learning_rate": 0.00017109244970662165,
|
| 905 |
+
"loss": 1.9446937561035156,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.11217391304347826,
|
| 910 |
+
"grad_norm": 0.04284852743148804,
|
| 911 |
+
"learning_rate": 0.0001710202415990805,
|
| 912 |
+
"loss": 1.8916801452636718,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.11304347826086956,
|
| 917 |
+
"grad_norm": 0.04538673162460327,
|
| 918 |
+
"learning_rate": 0.00017094738793494674,
|
| 919 |
+
"loss": 1.9826034545898437,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.11391304347826087,
|
| 924 |
+
"grad_norm": 0.045958247035741806,
|
| 925 |
+
"learning_rate": 0.0001708738892773313,
|
| 926 |
+
"loss": 1.918802261352539,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.11478260869565217,
|
| 931 |
+
"grad_norm": 0.043768592178821564,
|
| 932 |
+
"learning_rate": 0.00017079974619433054,
|
| 933 |
+
"loss": 1.9888408660888672,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.11565217391304349,
|
| 938 |
+
"grad_norm": 0.04695571959018707,
|
| 939 |
+
"learning_rate": 0.00017072495925902174,
|
| 940 |
+
"loss": 1.9092866897583007,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.11652173913043479,
|
| 945 |
+
"grad_norm": 0.04605986550450325,
|
| 946 |
+
"learning_rate": 0.00017064952904945885,
|
| 947 |
+
"loss": 1.9453262329101562,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.11739130434782609,
|
| 952 |
+
"grad_norm": 0.047600872814655304,
|
| 953 |
+
"learning_rate": 0.00017057345614866776,
|
| 954 |
+
"loss": 1.932131576538086,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.11826086956521739,
|
| 959 |
+
"grad_norm": 0.04641648009419441,
|
| 960 |
+
"learning_rate": 0.00017049674114464204,
|
| 961 |
+
"loss": 1.9330526351928712,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.1191304347826087,
|
| 966 |
+
"grad_norm": 0.04527211934328079,
|
| 967 |
+
"learning_rate": 0.0001704193846303382,
|
| 968 |
+
"loss": 1.9231037139892577,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.12,
|
| 973 |
+
"grad_norm": 0.049683600664138794,
|
| 974 |
+
"learning_rate": 0.0001703413872036714,
|
| 975 |
+
"loss": 1.988457489013672,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.1208695652173913,
|
| 980 |
+
"grad_norm": 0.045173171907663345,
|
| 981 |
+
"learning_rate": 0.0001702627494675104,
|
| 982 |
+
"loss": 1.952442741394043,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.12173913043478261,
|
| 987 |
+
"grad_norm": 0.04802786186337471,
|
| 988 |
+
"learning_rate": 0.00017018347202967326,
|
| 989 |
+
"loss": 1.915532112121582,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.12260869565217392,
|
| 994 |
+
"grad_norm": 0.04501411318778992,
|
| 995 |
+
"learning_rate": 0.0001701035555029225,
|
| 996 |
+
"loss": 1.9840251922607421,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.12347826086956522,
|
| 1001 |
+
"grad_norm": 0.043427255004644394,
|
| 1002 |
+
"learning_rate": 0.00017002300050496038,
|
| 1003 |
+
"loss": 1.8686183929443358,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.12434782608695652,
|
| 1008 |
+
"grad_norm": 0.04784533753991127,
|
| 1009 |
+
"learning_rate": 0.00016994180765842406,
|
| 1010 |
+
"loss": 1.8943784713745118,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.12521739130434784,
|
| 1015 |
+
"grad_norm": 0.044591404497623444,
|
| 1016 |
+
"learning_rate": 0.0001698599775908809,
|
| 1017 |
+
"loss": 1.8551948547363282,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.12608695652173912,
|
| 1022 |
+
"grad_norm": 0.04831645265221596,
|
| 1023 |
+
"learning_rate": 0.0001697775109348236,
|
| 1024 |
+
"loss": 1.890420913696289,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.12695652173913044,
|
| 1029 |
+
"grad_norm": 0.04690225422382355,
|
| 1030 |
+
"learning_rate": 0.0001696944083276652,
|
| 1031 |
+
"loss": 1.9546632766723633,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.12782608695652173,
|
| 1036 |
+
"grad_norm": 0.04488781467080116,
|
| 1037 |
+
"learning_rate": 0.00016961067041173432,
|
| 1038 |
+
"loss": 1.8585077285766602,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.12869565217391304,
|
| 1043 |
+
"grad_norm": 0.04684244468808174,
|
| 1044 |
+
"learning_rate": 0.00016952629783426992,
|
| 1045 |
+
"loss": 1.9355541229248048,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.12956521739130436,
|
| 1050 |
+
"grad_norm": 0.04604167491197586,
|
| 1051 |
+
"learning_rate": 0.00016944129124741664,
|
| 1052 |
+
"loss": 1.8615631103515624,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.13043478260869565,
|
| 1057 |
+
"grad_norm": 0.048263128846883774,
|
| 1058 |
+
"learning_rate": 0.00016935565130821956,
|
| 1059 |
+
"loss": 1.9315719604492188,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.13130434782608696,
|
| 1064 |
+
"grad_norm": 0.04678445681929588,
|
| 1065 |
+
"learning_rate": 0.0001692693786786191,
|
| 1066 |
+
"loss": 1.9562299728393555,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.13217391304347825,
|
| 1071 |
+
"grad_norm": 0.04416315257549286,
|
| 1072 |
+
"learning_rate": 0.0001691824740254461,
|
| 1073 |
+
"loss": 1.8920721054077148,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.13304347826086957,
|
| 1078 |
+
"grad_norm": 0.044212035834789276,
|
| 1079 |
+
"learning_rate": 0.00016909493802041628,
|
| 1080 |
+
"loss": 1.8729904174804688,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.13391304347826086,
|
| 1085 |
+
"grad_norm": 0.046913452446460724,
|
| 1086 |
+
"learning_rate": 0.0001690067713401256,
|
| 1087 |
+
"loss": 1.9575483322143554,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.13478260869565217,
|
| 1092 |
+
"grad_norm": 0.04588250815868378,
|
| 1093 |
+
"learning_rate": 0.00016891797466604444,
|
| 1094 |
+
"loss": 1.9156288146972655,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.1356521739130435,
|
| 1099 |
+
"grad_norm": 0.04520321637392044,
|
| 1100 |
+
"learning_rate": 0.00016882854868451285,
|
| 1101 |
+
"loss": 1.9655925750732421,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.13652173913043478,
|
| 1106 |
+
"grad_norm": 0.04560220241546631,
|
| 1107 |
+
"learning_rate": 0.00016873849408673486,
|
| 1108 |
+
"loss": 1.892664337158203,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.1373913043478261,
|
| 1113 |
+
"grad_norm": 0.04517337679862976,
|
| 1114 |
+
"learning_rate": 0.00016864781156877335,
|
| 1115 |
+
"loss": 1.9007608413696289,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.13826086956521738,
|
| 1120 |
+
"grad_norm": 0.046717483550310135,
|
| 1121 |
+
"learning_rate": 0.00016855650183154463,
|
| 1122 |
+
"loss": 1.9165136337280273,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.1391304347826087,
|
| 1127 |
+
"grad_norm": 0.04691331461071968,
|
| 1128 |
+
"learning_rate": 0.0001684645655808129,
|
| 1129 |
+
"loss": 1.8728553771972656,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.14,
|
| 1134 |
+
"grad_norm": 0.04566628113389015,
|
| 1135 |
+
"learning_rate": 0.00016837200352718507,
|
| 1136 |
+
"loss": 1.883155059814453,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.1408695652173913,
|
| 1141 |
+
"grad_norm": 0.04991758242249489,
|
| 1142 |
+
"learning_rate": 0.00016827881638610496,
|
| 1143 |
+
"loss": 1.9257890701293945,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.14173913043478262,
|
| 1148 |
+
"grad_norm": 0.04652441665530205,
|
| 1149 |
+
"learning_rate": 0.0001681850048778479,
|
| 1150 |
+
"loss": 1.9192398071289063,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.1426086956521739,
|
| 1155 |
+
"grad_norm": 0.04636216163635254,
|
| 1156 |
+
"learning_rate": 0.00016809056972751527,
|
| 1157 |
+
"loss": 1.9711288452148437,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.14347826086956522,
|
| 1162 |
+
"grad_norm": 0.04579229652881622,
|
| 1163 |
+
"learning_rate": 0.00016799551166502873,
|
| 1164 |
+
"loss": 1.9829723358154296,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.1443478260869565,
|
| 1169 |
+
"grad_norm": 0.044148631393909454,
|
| 1170 |
+
"learning_rate": 0.00016789983142512463,
|
| 1171 |
+
"loss": 1.912728500366211,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.14521739130434783,
|
| 1176 |
+
"grad_norm": 0.046788472682237625,
|
| 1177 |
+
"learning_rate": 0.00016780352974734834,
|
| 1178 |
+
"loss": 1.938332748413086,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.14608695652173914,
|
| 1183 |
+
"grad_norm": 0.04814344644546509,
|
| 1184 |
+
"learning_rate": 0.00016770660737604862,
|
| 1185 |
+
"loss": 1.9140899658203125,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.14695652173913043,
|
| 1190 |
+
"grad_norm": 0.046944644302129745,
|
| 1191 |
+
"learning_rate": 0.00016760906506037166,
|
| 1192 |
+
"loss": 1.9023950576782227,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.14782608695652175,
|
| 1197 |
+
"grad_norm": 0.04715333133935928,
|
| 1198 |
+
"learning_rate": 0.00016751090355425548,
|
| 1199 |
+
"loss": 1.9201131820678712,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.14869565217391303,
|
| 1204 |
+
"grad_norm": 0.045025281608104706,
|
| 1205 |
+
"learning_rate": 0.00016741212361642407,
|
| 1206 |
+
"loss": 1.885734748840332,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.14956521739130435,
|
| 1211 |
+
"grad_norm": 0.04626510292291641,
|
| 1212 |
+
"learning_rate": 0.00016731272601038137,
|
| 1213 |
+
"loss": 1.939837646484375,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.15043478260869564,
|
| 1218 |
+
"grad_norm": 0.04652830585837364,
|
| 1219 |
+
"learning_rate": 0.00016721271150440568,
|
| 1220 |
+
"loss": 1.880080223083496,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.15130434782608695,
|
| 1225 |
+
"grad_norm": 0.044095490127801895,
|
| 1226 |
+
"learning_rate": 0.00016711208087154334,
|
| 1227 |
+
"loss": 1.9228801727294922,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.15217391304347827,
|
| 1232 |
+
"grad_norm": 0.04645839333534241,
|
| 1233 |
+
"learning_rate": 0.00016701083488960302,
|
| 1234 |
+
"loss": 1.939291763305664,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.15304347826086956,
|
| 1239 |
+
"grad_norm": 0.04704873263835907,
|
| 1240 |
+
"learning_rate": 0.00016690897434114965,
|
| 1241 |
+
"loss": 1.9802566528320313,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.15391304347826087,
|
| 1246 |
+
"grad_norm": 0.044596049934625626,
|
| 1247 |
+
"learning_rate": 0.00016680650001349838,
|
| 1248 |
+
"loss": 1.9283245086669922,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.15478260869565216,
|
| 1253 |
+
"grad_norm": 0.04547381401062012,
|
| 1254 |
+
"learning_rate": 0.0001667034126987084,
|
| 1255 |
+
"loss": 1.882451629638672,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.15565217391304348,
|
| 1260 |
+
"grad_norm": 0.04559491574764252,
|
| 1261 |
+
"learning_rate": 0.00016659971319357694,
|
| 1262 |
+
"loss": 1.9023427963256836,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.1565217391304348,
|
| 1267 |
+
"grad_norm": 0.04723436385393143,
|
| 1268 |
+
"learning_rate": 0.00016649540229963304,
|
| 1269 |
+
"loss": 1.9267019271850585,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.15739130434782608,
|
| 1274 |
+
"grad_norm": 0.04553121328353882,
|
| 1275 |
+
"learning_rate": 0.00016639048082313138,
|
| 1276 |
+
"loss": 1.8939033508300782,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.1582608695652174,
|
| 1281 |
+
"grad_norm": 0.046808917075395584,
|
| 1282 |
+
"learning_rate": 0.000166284949575046,
|
| 1283 |
+
"loss": 1.8823970794677733,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 0.1591304347826087,
|
| 1288 |
+
"grad_norm": 0.04732776805758476,
|
| 1289 |
+
"learning_rate": 0.00016617880937106413,
|
| 1290 |
+
"loss": 1.8905269622802734,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 0.16,
|
| 1295 |
+
"grad_norm": 0.04512569308280945,
|
| 1296 |
+
"learning_rate": 0.00016607206103157982,
|
| 1297 |
+
"loss": 1.891006851196289,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.1608695652173913,
|
| 1302 |
+
"grad_norm": 0.0452556274831295,
|
| 1303 |
+
"learning_rate": 0.0001659647053816876,
|
| 1304 |
+
"loss": 1.8488157272338868,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.1617391304347826,
|
| 1309 |
+
"grad_norm": 0.04695947468280792,
|
| 1310 |
+
"learning_rate": 0.00016585674325117613,
|
| 1311 |
+
"loss": 1.8792579650878907,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 0.16260869565217392,
|
| 1316 |
+
"grad_norm": 0.046590715646743774,
|
| 1317 |
+
"learning_rate": 0.0001657481754745217,
|
| 1318 |
+
"loss": 1.905181884765625,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 0.1634782608695652,
|
| 1323 |
+
"grad_norm": 0.04671342298388481,
|
| 1324 |
+
"learning_rate": 0.00016563900289088187,
|
| 1325 |
+
"loss": 1.879676055908203,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 0.16434782608695653,
|
| 1330 |
+
"grad_norm": 0.04555339738726616,
|
| 1331 |
+
"learning_rate": 0.0001655292263440891,
|
| 1332 |
+
"loss": 1.9276203155517577,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 0.16521739130434782,
|
| 1337 |
+
"grad_norm": 0.04350179061293602,
|
| 1338 |
+
"learning_rate": 0.00016541884668264383,
|
| 1339 |
+
"loss": 1.8957508087158204,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.16608695652173913,
|
| 1344 |
+
"grad_norm": 0.04489278420805931,
|
| 1345 |
+
"learning_rate": 0.0001653078647597084,
|
| 1346 |
+
"loss": 1.9363662719726562,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.16695652173913045,
|
| 1351 |
+
"grad_norm": 0.04279284551739693,
|
| 1352 |
+
"learning_rate": 0.00016519628143310015,
|
| 1353 |
+
"loss": 1.953699493408203,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 0.16782608695652174,
|
| 1358 |
+
"grad_norm": 0.04596225544810295,
|
| 1359 |
+
"learning_rate": 0.00016508409756528486,
|
| 1360 |
+
"loss": 1.9585641860961913,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 0.16869565217391305,
|
| 1365 |
+
"grad_norm": 0.04413312301039696,
|
| 1366 |
+
"learning_rate": 0.00016497131402337016,
|
| 1367 |
+
"loss": 1.8821155548095703,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 0.16956521739130434,
|
| 1372 |
+
"grad_norm": 0.04498974233865738,
|
| 1373 |
+
"learning_rate": 0.00016485793167909864,
|
| 1374 |
+
"loss": 1.9019199371337892,
|
| 1375 |
+
"step": 1950
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 0.17043478260869566,
|
| 1379 |
+
"grad_norm": 0.04706019163131714,
|
| 1380 |
+
"learning_rate": 0.00016474395140884143,
|
| 1381 |
+
"loss": 1.8370037078857422,
|
| 1382 |
+
"step": 1960
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.17130434782608694,
|
| 1386 |
+
"grad_norm": 0.04894762858748436,
|
| 1387 |
+
"learning_rate": 0.0001646293740935911,
|
| 1388 |
+
"loss": 1.9752735137939452,
|
| 1389 |
+
"step": 1970
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.17217391304347826,
|
| 1393 |
+
"grad_norm": 0.04635745286941528,
|
| 1394 |
+
"learning_rate": 0.00016451420061895498,
|
| 1395 |
+
"loss": 1.9637317657470703,
|
| 1396 |
+
"step": 1980
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 0.17304347826086958,
|
| 1400 |
+
"grad_norm": 0.04573964327573776,
|
| 1401 |
+
"learning_rate": 0.00016439843187514836,
|
| 1402 |
+
"loss": 1.9148262023925782,
|
| 1403 |
+
"step": 1990
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 0.17391304347826086,
|
| 1407 |
+
"grad_norm": 0.045651230961084366,
|
| 1408 |
+
"learning_rate": 0.00016428206875698762,
|
| 1409 |
+
"loss": 1.8789260864257813,
|
| 1410 |
+
"step": 2000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 0.17478260869565218,
|
| 1414 |
+
"grad_norm": 0.047087036073207855,
|
| 1415 |
+
"learning_rate": 0.00016416511216388316,
|
| 1416 |
+
"loss": 1.913290786743164,
|
| 1417 |
+
"step": 2010
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 0.17565217391304347,
|
| 1421 |
+
"grad_norm": 0.045222099870443344,
|
| 1422 |
+
"learning_rate": 0.0001640475629998326,
|
| 1423 |
+
"loss": 1.9029350280761719,
|
| 1424 |
+
"step": 2020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 0.17652173913043478,
|
| 1428 |
+
"grad_norm": 0.04745696485042572,
|
| 1429 |
+
"learning_rate": 0.00016392942217341382,
|
| 1430 |
+
"loss": 1.9319425582885743,
|
| 1431 |
+
"step": 2030
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 0.17739130434782607,
|
| 1435 |
+
"grad_norm": 0.047783516347408295,
|
| 1436 |
+
"learning_rate": 0.00016381069059777769,
|
| 1437 |
+
"loss": 1.8861225128173829,
|
| 1438 |
+
"step": 2040
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 0.1782608695652174,
|
| 1442 |
+
"grad_norm": 0.04685746133327484,
|
| 1443 |
+
"learning_rate": 0.00016369136919064132,
|
| 1444 |
+
"loss": 1.990982437133789,
|
| 1445 |
+
"step": 2050
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 0.1791304347826087,
|
| 1449 |
+
"grad_norm": 0.0454823300242424,
|
| 1450 |
+
"learning_rate": 0.0001635714588742808,
|
| 1451 |
+
"loss": 1.859075164794922,
|
| 1452 |
+
"step": 2060
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.18,
|
| 1456 |
+
"grad_norm": 0.044694554060697556,
|
| 1457 |
+
"learning_rate": 0.00016345096057552409,
|
| 1458 |
+
"loss": 1.926215934753418,
|
| 1459 |
+
"step": 2070
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 0.1808695652173913,
|
| 1463 |
+
"grad_norm": 0.04416250064969063,
|
| 1464 |
+
"learning_rate": 0.00016332987522574383,
|
| 1465 |
+
"loss": 1.8868453979492188,
|
| 1466 |
+
"step": 2080
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.1817391304347826,
|
| 1470 |
+
"grad_norm": 0.04497243091464043,
|
| 1471 |
+
"learning_rate": 0.00016320820376085026,
|
| 1472 |
+
"loss": 1.902182388305664,
|
| 1473 |
+
"step": 2090
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.1826086956521739,
|
| 1477 |
+
"grad_norm": 0.0451941043138504,
|
| 1478 |
+
"learning_rate": 0.00016308594712128384,
|
| 1479 |
+
"loss": 1.960106658935547,
|
| 1480 |
+
"step": 2100
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 0.18347826086956523,
|
| 1484 |
+
"grad_norm": 0.046650130301713943,
|
| 1485 |
+
"learning_rate": 0.0001629631062520081,
|
| 1486 |
+
"loss": 1.9553258895874024,
|
| 1487 |
+
"step": 2110
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 0.18434782608695652,
|
| 1491 |
+
"grad_norm": 0.04722883552312851,
|
| 1492 |
+
"learning_rate": 0.0001628396821025022,
|
| 1493 |
+
"loss": 1.9351499557495118,
|
| 1494 |
+
"step": 2120
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 0.18521739130434783,
|
| 1498 |
+
"grad_norm": 0.04475412890315056,
|
| 1499 |
+
"learning_rate": 0.00016271567562675378,
|
| 1500 |
+
"loss": 1.8991554260253907,
|
| 1501 |
+
"step": 2130
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 0.18608695652173912,
|
| 1505 |
+
"grad_norm": 0.0482119657099247,
|
| 1506 |
+
"learning_rate": 0.0001625910877832514,
|
| 1507 |
+
"loss": 1.9038169860839844,
|
| 1508 |
+
"step": 2140
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 0.18695652173913044,
|
| 1512 |
+
"grad_norm": 0.04647846147418022,
|
| 1513 |
+
"learning_rate": 0.0001624659195349772,
|
| 1514 |
+
"loss": 1.8781488418579102,
|
| 1515 |
+
"step": 2150
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 0.18782608695652173,
|
| 1519 |
+
"grad_norm": 0.04708510637283325,
|
| 1520 |
+
"learning_rate": 0.00016234017184939953,
|
| 1521 |
+
"loss": 1.951561164855957,
|
| 1522 |
+
"step": 2160
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 0.18869565217391304,
|
| 1526 |
+
"grad_norm": 0.04458949714899063,
|
| 1527 |
+
"learning_rate": 0.00016221384569846539,
|
| 1528 |
+
"loss": 1.9635744094848633,
|
| 1529 |
+
"step": 2170
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 0.18956521739130436,
|
| 1533 |
+
"grad_norm": 0.04679902642965317,
|
| 1534 |
+
"learning_rate": 0.00016208694205859287,
|
| 1535 |
+
"loss": 1.941354751586914,
|
| 1536 |
+
"step": 2180
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 0.19043478260869565,
|
| 1540 |
+
"grad_norm": 0.04542798921465874,
|
| 1541 |
+
"learning_rate": 0.0001619594619106638,
|
| 1542 |
+
"loss": 1.9266124725341798,
|
| 1543 |
+
"step": 2190
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 0.19130434782608696,
|
| 1547 |
+
"grad_norm": 0.04602201655507088,
|
| 1548 |
+
"learning_rate": 0.00016183140624001594,
|
| 1549 |
+
"loss": 1.968562126159668,
|
| 1550 |
+
"step": 2200
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 0.19217391304347825,
|
| 1554 |
+
"grad_norm": 0.045000337064266205,
|
| 1555 |
+
"learning_rate": 0.0001617027760364355,
|
| 1556 |
+
"loss": 1.9628772735595703,
|
| 1557 |
+
"step": 2210
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 0.19304347826086957,
|
| 1561 |
+
"grad_norm": 0.043177004903554916,
|
| 1562 |
+
"learning_rate": 0.00016157357229414941,
|
| 1563 |
+
"loss": 1.918726348876953,
|
| 1564 |
+
"step": 2220
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 0.19391304347826088,
|
| 1568 |
+
"grad_norm": 0.045546501874923706,
|
| 1569 |
+
"learning_rate": 0.00016144379601181773,
|
| 1570 |
+
"loss": 1.8972591400146483,
|
| 1571 |
+
"step": 2230
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 0.19478260869565217,
|
| 1575 |
+
"grad_norm": 0.04712609574198723,
|
| 1576 |
+
"learning_rate": 0.00016131344819252585,
|
| 1577 |
+
"loss": 1.8813795089721679,
|
| 1578 |
+
"step": 2240
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 0.1956521739130435,
|
| 1582 |
+
"grad_norm": 0.04688692092895508,
|
| 1583 |
+
"learning_rate": 0.00016118252984377674,
|
| 1584 |
+
"loss": 1.9304527282714843,
|
| 1585 |
+
"step": 2250
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 0.19652173913043477,
|
| 1589 |
+
"grad_norm": 0.04640372470021248,
|
| 1590 |
+
"learning_rate": 0.00016105104197748328,
|
| 1591 |
+
"loss": 1.9001270294189454,
|
| 1592 |
+
"step": 2260
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 0.1973913043478261,
|
| 1596 |
+
"grad_norm": 0.04748000577092171,
|
| 1597 |
+
"learning_rate": 0.00016091898560996026,
|
| 1598 |
+
"loss": 1.8875625610351563,
|
| 1599 |
+
"step": 2270
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 0.19826086956521738,
|
| 1603 |
+
"grad_norm": 0.04630134627223015,
|
| 1604 |
+
"learning_rate": 0.00016078636176191662,
|
| 1605 |
+
"loss": 1.880694580078125,
|
| 1606 |
+
"step": 2280
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 0.1991304347826087,
|
| 1610 |
+
"grad_norm": 0.04616926237940788,
|
| 1611 |
+
"learning_rate": 0.00016065317145844756,
|
| 1612 |
+
"loss": 1.919076156616211,
|
| 1613 |
+
"step": 2290
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 0.2,
|
| 1617 |
+
"grad_norm": 0.04642150551080704,
|
| 1618 |
+
"learning_rate": 0.00016051941572902664,
|
| 1619 |
+
"loss": 1.9145648956298829,
|
| 1620 |
+
"step": 2300
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 0.2008695652173913,
|
| 1624 |
+
"grad_norm": 0.04683443531394005,
|
| 1625 |
+
"learning_rate": 0.00016038509560749778,
|
| 1626 |
+
"loss": 1.998061180114746,
|
| 1627 |
+
"step": 2310
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 0.20173913043478262,
|
| 1631 |
+
"grad_norm": 0.04580046609044075,
|
| 1632 |
+
"learning_rate": 0.00016025021213206722,
|
| 1633 |
+
"loss": 1.8663105010986327,
|
| 1634 |
+
"step": 2320
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.2026086956521739,
|
| 1638 |
+
"grad_norm": 0.046270985156297684,
|
| 1639 |
+
"learning_rate": 0.00016011476634529567,
|
| 1640 |
+
"loss": 1.9407358169555664,
|
| 1641 |
+
"step": 2330
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 0.20347826086956522,
|
| 1645 |
+
"grad_norm": 0.047260086983442307,
|
| 1646 |
+
"learning_rate": 0.00015997875929409005,
|
| 1647 |
+
"loss": 1.8749610900878906,
|
| 1648 |
+
"step": 2340
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 0.20434782608695654,
|
| 1652 |
+
"grad_norm": 0.04815071448683739,
|
| 1653 |
+
"learning_rate": 0.00015984219202969544,
|
| 1654 |
+
"loss": 1.9252590179443358,
|
| 1655 |
+
"step": 2350
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.20521739130434782,
|
| 1659 |
+
"grad_norm": 0.04547717794775963,
|
| 1660 |
+
"learning_rate": 0.00015970506560768715,
|
| 1661 |
+
"loss": 1.8544036865234375,
|
| 1662 |
+
"step": 2360
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 0.20608695652173914,
|
| 1666 |
+
"grad_norm": 0.04534965381026268,
|
| 1667 |
+
"learning_rate": 0.00015956738108796227,
|
| 1668 |
+
"loss": 1.9171245574951172,
|
| 1669 |
+
"step": 2370
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 0.20695652173913043,
|
| 1673 |
+
"grad_norm": 0.0445626862347126,
|
| 1674 |
+
"learning_rate": 0.00015942913953473172,
|
| 1675 |
+
"loss": 1.8566919326782227,
|
| 1676 |
+
"step": 2380
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.20782608695652174,
|
| 1680 |
+
"grad_norm": 0.04675230383872986,
|
| 1681 |
+
"learning_rate": 0.00015929034201651185,
|
| 1682 |
+
"loss": 1.9092636108398438,
|
| 1683 |
+
"step": 2390
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.20869565217391303,
|
| 1687 |
+
"grad_norm": 0.04733717441558838,
|
| 1688 |
+
"learning_rate": 0.00015915098960611628,
|
| 1689 |
+
"loss": 1.9153020858764649,
|
| 1690 |
+
"step": 2400
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.20956521739130435,
|
| 1694 |
+
"grad_norm": 0.04689701274037361,
|
| 1695 |
+
"learning_rate": 0.0001590110833806476,
|
| 1696 |
+
"loss": 1.8826875686645508,
|
| 1697 |
+
"step": 2410
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.21043478260869566,
|
| 1701 |
+
"grad_norm": 0.04545078054070473,
|
| 1702 |
+
"learning_rate": 0.000158870624421489,
|
| 1703 |
+
"loss": 1.8986186981201172,
|
| 1704 |
+
"step": 2420
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 0.21130434782608695,
|
| 1708 |
+
"grad_norm": 0.04620114341378212,
|
| 1709 |
+
"learning_rate": 0.00015872961381429592,
|
| 1710 |
+
"loss": 1.9260257720947265,
|
| 1711 |
+
"step": 2430
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 0.21217391304347827,
|
| 1715 |
+
"grad_norm": 0.047381263226270676,
|
| 1716 |
+
"learning_rate": 0.00015858805264898777,
|
| 1717 |
+
"loss": 1.928044319152832,
|
| 1718 |
+
"step": 2440
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.21304347826086956,
|
| 1722 |
+
"grad_norm": 0.04613082483410835,
|
| 1723 |
+
"learning_rate": 0.00015844594201973935,
|
| 1724 |
+
"loss": 1.9063533782958983,
|
| 1725 |
+
"step": 2450
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.21391304347826087,
|
| 1729 |
+
"grad_norm": 0.04579924792051315,
|
| 1730 |
+
"learning_rate": 0.00015830328302497235,
|
| 1731 |
+
"loss": 1.8655834197998047,
|
| 1732 |
+
"step": 2460
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 0.21478260869565216,
|
| 1736 |
+
"grad_norm": 0.04783694073557854,
|
| 1737 |
+
"learning_rate": 0.0001581600767673472,
|
| 1738 |
+
"loss": 1.8869522094726563,
|
| 1739 |
+
"step": 2470
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.21565217391304348,
|
| 1743 |
+
"grad_norm": 0.04724857956171036,
|
| 1744 |
+
"learning_rate": 0.0001580163243537541,
|
| 1745 |
+
"loss": 1.911419677734375,
|
| 1746 |
+
"step": 2480
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 0.2165217391304348,
|
| 1750 |
+
"grad_norm": 0.046219322830438614,
|
| 1751 |
+
"learning_rate": 0.00015787202689530476,
|
| 1752 |
+
"loss": 1.878607940673828,
|
| 1753 |
+
"step": 2490
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.21739130434782608,
|
| 1757 |
+
"grad_norm": 0.04494387283921242,
|
| 1758 |
+
"learning_rate": 0.00015772718550732375,
|
| 1759 |
+
"loss": 1.9234241485595702,
|
| 1760 |
+
"step": 2500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 0.2182608695652174,
|
| 1764 |
+
"grad_norm": 0.04502217471599579,
|
| 1765 |
+
"learning_rate": 0.00015758180130933985,
|
| 1766 |
+
"loss": 1.915613555908203,
|
| 1767 |
+
"step": 2510
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 0.21913043478260869,
|
| 1771 |
+
"grad_norm": 0.04806201159954071,
|
| 1772 |
+
"learning_rate": 0.00015743587542507736,
|
| 1773 |
+
"loss": 1.8788349151611328,
|
| 1774 |
+
"step": 2520
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 0.22,
|
| 1778 |
+
"grad_norm": 0.04505891352891922,
|
| 1779 |
+
"learning_rate": 0.00015728940898244748,
|
| 1780 |
+
"loss": 1.892514419555664,
|
| 1781 |
+
"step": 2530
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 0.22086956521739132,
|
| 1785 |
+
"grad_norm": 0.047691989690065384,
|
| 1786 |
+
"learning_rate": 0.00015714240311353963,
|
| 1787 |
+
"loss": 1.919693374633789,
|
| 1788 |
+
"step": 2540
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 0.2217391304347826,
|
| 1792 |
+
"grad_norm": 0.04654225334525108,
|
| 1793 |
+
"learning_rate": 0.00015699485895461252,
|
| 1794 |
+
"loss": 1.8307640075683593,
|
| 1795 |
+
"step": 2550
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 0.22260869565217392,
|
| 1799 |
+
"grad_norm": 0.04682806134223938,
|
| 1800 |
+
"learning_rate": 0.00015684677764608564,
|
| 1801 |
+
"loss": 1.8828701019287108,
|
| 1802 |
+
"step": 2560
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 0.2234782608695652,
|
| 1806 |
+
"grad_norm": 0.04656338319182396,
|
| 1807 |
+
"learning_rate": 0.00015669816033253014,
|
| 1808 |
+
"loss": 1.877579689025879,
|
| 1809 |
+
"step": 2570
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 0.22434782608695653,
|
| 1813 |
+
"grad_norm": 0.0479404553771019,
|
| 1814 |
+
"learning_rate": 0.00015654900816266024,
|
| 1815 |
+
"loss": 1.894699478149414,
|
| 1816 |
+
"step": 2580
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 0.2252173913043478,
|
| 1820 |
+
"grad_norm": 0.047481246292591095,
|
| 1821 |
+
"learning_rate": 0.00015639932228932418,
|
| 1822 |
+
"loss": 1.8986604690551758,
|
| 1823 |
+
"step": 2590
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 0.22608695652173913,
|
| 1827 |
+
"grad_norm": 0.04713468998670578,
|
| 1828 |
+
"learning_rate": 0.0001562491038694955,
|
| 1829 |
+
"loss": 1.9140901565551758,
|
| 1830 |
+
"step": 2600
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 0.22695652173913045,
|
| 1834 |
+
"grad_norm": 0.046244435012340546,
|
| 1835 |
+
"learning_rate": 0.0001560983540642638,
|
| 1836 |
+
"loss": 1.8460613250732423,
|
| 1837 |
+
"step": 2610
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 0.22782608695652173,
|
| 1841 |
+
"grad_norm": 0.0466030091047287,
|
| 1842 |
+
"learning_rate": 0.00015594707403882607,
|
| 1843 |
+
"loss": 1.9079011917114257,
|
| 1844 |
+
"step": 2620
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 0.22869565217391305,
|
| 1848 |
+
"grad_norm": 0.04684355854988098,
|
| 1849 |
+
"learning_rate": 0.00015579526496247752,
|
| 1850 |
+
"loss": 1.923720932006836,
|
| 1851 |
+
"step": 2630
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 0.22956521739130434,
|
| 1855 |
+
"grad_norm": 0.04607565701007843,
|
| 1856 |
+
"learning_rate": 0.00015564292800860252,
|
| 1857 |
+
"loss": 1.8996570587158204,
|
| 1858 |
+
"step": 2640
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 0.23043478260869565,
|
| 1862 |
+
"grad_norm": 0.0482359416782856,
|
| 1863 |
+
"learning_rate": 0.00015549006435466566,
|
| 1864 |
+
"loss": 1.8546255111694336,
|
| 1865 |
+
"step": 2650
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 0.23130434782608697,
|
| 1869 |
+
"grad_norm": 0.04718746244907379,
|
| 1870 |
+
"learning_rate": 0.0001553366751822026,
|
| 1871 |
+
"loss": 1.9036752700805664,
|
| 1872 |
+
"step": 2660
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 0.23217391304347826,
|
| 1876 |
+
"grad_norm": 0.04525565356016159,
|
| 1877 |
+
"learning_rate": 0.00015518276167681078,
|
| 1878 |
+
"loss": 1.878532600402832,
|
| 1879 |
+
"step": 2670
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 0.23304347826086957,
|
| 1883 |
+
"grad_norm": 0.045305028557777405,
|
| 1884 |
+
"learning_rate": 0.0001550283250281405,
|
| 1885 |
+
"loss": 1.8548662185668945,
|
| 1886 |
+
"step": 2680
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 0.23391304347826086,
|
| 1890 |
+
"grad_norm": 0.04721986874938011,
|
| 1891 |
+
"learning_rate": 0.00015487336642988563,
|
| 1892 |
+
"loss": 1.898862075805664,
|
| 1893 |
+
"step": 2690
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 0.23478260869565218,
|
| 1897 |
+
"grad_norm": 0.0476800911128521,
|
| 1898 |
+
"learning_rate": 0.00015471788707977424,
|
| 1899 |
+
"loss": 1.8859397888183593,
|
| 1900 |
+
"step": 2700
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 0.23565217391304347,
|
| 1904 |
+
"grad_norm": 0.04801041632890701,
|
| 1905 |
+
"learning_rate": 0.00015456188817955963,
|
| 1906 |
+
"loss": 1.884882926940918,
|
| 1907 |
+
"step": 2710
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 0.23652173913043478,
|
| 1911 |
+
"grad_norm": 0.04432157799601555,
|
| 1912 |
+
"learning_rate": 0.00015440537093501077,
|
| 1913 |
+
"loss": 1.9631668090820313,
|
| 1914 |
+
"step": 2720
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 0.2373913043478261,
|
| 1918 |
+
"grad_norm": 0.045344945043325424,
|
| 1919 |
+
"learning_rate": 0.00015424833655590313,
|
| 1920 |
+
"loss": 1.9439357757568358,
|
| 1921 |
+
"step": 2730
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 0.2382608695652174,
|
| 1925 |
+
"grad_norm": 0.046976640820503235,
|
| 1926 |
+
"learning_rate": 0.00015409078625600926,
|
| 1927 |
+
"loss": 1.8975433349609374,
|
| 1928 |
+
"step": 2740
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 0.2391304347826087,
|
| 1932 |
+
"grad_norm": 0.04438307136297226,
|
| 1933 |
+
"learning_rate": 0.0001539327212530895,
|
| 1934 |
+
"loss": 1.870760154724121,
|
| 1935 |
+
"step": 2750
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 0.24,
|
| 1939 |
+
"grad_norm": 0.04586169123649597,
|
| 1940 |
+
"learning_rate": 0.00015377414276888246,
|
| 1941 |
+
"loss": 1.8693674087524415,
|
| 1942 |
+
"step": 2760
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 0.2408695652173913,
|
| 1946 |
+
"grad_norm": 0.04416853189468384,
|
| 1947 |
+
"learning_rate": 0.00015361505202909564,
|
| 1948 |
+
"loss": 1.8614776611328125,
|
| 1949 |
+
"step": 2770
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 0.2417391304347826,
|
| 1953 |
+
"grad_norm": 0.04704264551401138,
|
| 1954 |
+
"learning_rate": 0.0001534554502633959,
|
| 1955 |
+
"loss": 1.9111223220825195,
|
| 1956 |
+
"step": 2780
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 0.2426086956521739,
|
| 1960 |
+
"grad_norm": 0.04835261031985283,
|
| 1961 |
+
"learning_rate": 0.00015329533870540003,
|
| 1962 |
+
"loss": 1.9522029876708984,
|
| 1963 |
+
"step": 2790
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 0.24347826086956523,
|
| 1967 |
+
"grad_norm": 0.048007503151893616,
|
| 1968 |
+
"learning_rate": 0.00015313471859266522,
|
| 1969 |
+
"loss": 1.9032674789428712,
|
| 1970 |
+
"step": 2800
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 0.24434782608695652,
|
| 1974 |
+
"grad_norm": 0.04623431712388992,
|
| 1975 |
+
"learning_rate": 0.00015297359116667934,
|
| 1976 |
+
"loss": 1.911952018737793,
|
| 1977 |
+
"step": 2810
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 0.24521739130434783,
|
| 1981 |
+
"grad_norm": 0.045389365404844284,
|
| 1982 |
+
"learning_rate": 0.00015281195767285162,
|
| 1983 |
+
"loss": 1.945793914794922,
|
| 1984 |
+
"step": 2820
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 0.24608695652173912,
|
| 1988 |
+
"grad_norm": 0.046615757048130035,
|
| 1989 |
+
"learning_rate": 0.00015264981936050269,
|
| 1990 |
+
"loss": 1.8489097595214843,
|
| 1991 |
+
"step": 2830
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 0.24695652173913044,
|
| 1995 |
+
"grad_norm": 0.04565439373254776,
|
| 1996 |
+
"learning_rate": 0.00015248717748285521,
|
| 1997 |
+
"loss": 1.7807727813720704,
|
| 1998 |
+
"step": 2840
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 0.24782608695652175,
|
| 2002 |
+
"grad_norm": 0.04634851589798927,
|
| 2003 |
+
"learning_rate": 0.00015232403329702406,
|
| 2004 |
+
"loss": 1.8626737594604492,
|
| 2005 |
+
"step": 2850
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 0.24869565217391304,
|
| 2009 |
+
"grad_norm": 0.043341245502233505,
|
| 2010 |
+
"learning_rate": 0.00015216038806400655,
|
| 2011 |
+
"loss": 1.885053825378418,
|
| 2012 |
+
"step": 2860
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 0.24956521739130436,
|
| 2016 |
+
"grad_norm": 0.044841159135103226,
|
| 2017 |
+
"learning_rate": 0.00015199624304867285,
|
| 2018 |
+
"loss": 1.877364158630371,
|
| 2019 |
+
"step": 2870
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 0.25043478260869567,
|
| 2023 |
+
"grad_norm": 0.045520294457674026,
|
| 2024 |
+
"learning_rate": 0.00015183159951975605,
|
| 2025 |
+
"loss": 1.896337127685547,
|
| 2026 |
+
"step": 2880
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 0.25130434782608696,
|
| 2030 |
+
"grad_norm": 0.04885012283921242,
|
| 2031 |
+
"learning_rate": 0.00015166645874984248,
|
| 2032 |
+
"loss": 1.9359861373901368,
|
| 2033 |
+
"step": 2890
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 0.25217391304347825,
|
| 2037 |
+
"grad_norm": 0.04632116109132767,
|
| 2038 |
+
"learning_rate": 0.0001515008220153618,
|
| 2039 |
+
"loss": 1.8857734680175782,
|
| 2040 |
+
"step": 2900
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 0.2530434782608696,
|
| 2044 |
+
"grad_norm": 0.047859784215688705,
|
| 2045 |
+
"learning_rate": 0.0001513346905965771,
|
| 2046 |
+
"loss": 1.8554449081420898,
|
| 2047 |
+
"step": 2910
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 0.2539130434782609,
|
| 2051 |
+
"grad_norm": 0.046239182353019714,
|
| 2052 |
+
"learning_rate": 0.00015116806577757512,
|
| 2053 |
+
"loss": 1.7961738586425782,
|
| 2054 |
+
"step": 2920
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 0.25478260869565217,
|
| 2058 |
+
"grad_norm": 0.04520752280950546,
|
| 2059 |
+
"learning_rate": 0.00015100094884625624,
|
| 2060 |
+
"loss": 1.8724737167358398,
|
| 2061 |
+
"step": 2930
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 0.25565217391304346,
|
| 2065 |
+
"grad_norm": 0.046254873275756836,
|
| 2066 |
+
"learning_rate": 0.00015083334109432454,
|
| 2067 |
+
"loss": 1.8898199081420899,
|
| 2068 |
+
"step": 2940
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 0.2565217391304348,
|
| 2072 |
+
"grad_norm": 0.04747215658426285,
|
| 2073 |
+
"learning_rate": 0.00015066524381727778,
|
| 2074 |
+
"loss": 1.8878746032714844,
|
| 2075 |
+
"step": 2950
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 0.2573913043478261,
|
| 2079 |
+
"grad_norm": 0.048975683748722076,
|
| 2080 |
+
"learning_rate": 0.00015049665831439753,
|
| 2081 |
+
"loss": 1.8283885955810546,
|
| 2082 |
+
"step": 2960
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 0.2582608695652174,
|
| 2086 |
+
"grad_norm": 0.04566105827689171,
|
| 2087 |
+
"learning_rate": 0.0001503275858887389,
|
| 2088 |
+
"loss": 1.840024185180664,
|
| 2089 |
+
"step": 2970
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 0.2591304347826087,
|
| 2093 |
+
"grad_norm": 0.046260230243206024,
|
| 2094 |
+
"learning_rate": 0.0001501580278471207,
|
| 2095 |
+
"loss": 1.867376708984375,
|
| 2096 |
+
"step": 2980
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 0.26,
|
| 2100 |
+
"grad_norm": 0.046284548938274384,
|
| 2101 |
+
"learning_rate": 0.00014998798550011516,
|
| 2102 |
+
"loss": 1.886456298828125,
|
| 2103 |
+
"step": 2990
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 0.2608695652173913,
|
| 2107 |
+
"grad_norm": 0.046349357813596725,
|
| 2108 |
+
"learning_rate": 0.00014981746016203793,
|
| 2109 |
+
"loss": 1.8599205017089844,
|
| 2110 |
+
"step": 3000
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 0.2617391304347826,
|
| 2114 |
+
"grad_norm": 0.046956196427345276,
|
| 2115 |
+
"learning_rate": 0.0001496464531509378,
|
| 2116 |
+
"loss": 1.8856639862060547,
|
| 2117 |
+
"step": 3010
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 0.26260869565217393,
|
| 2121 |
+
"grad_norm": 0.045312799513339996,
|
| 2122 |
+
"learning_rate": 0.00014947496578858666,
|
| 2123 |
+
"loss": 1.8629747390747071,
|
| 2124 |
+
"step": 3020
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 0.2634782608695652,
|
| 2128 |
+
"grad_norm": 0.04610862210392952,
|
| 2129 |
+
"learning_rate": 0.00014930299940046907,
|
| 2130 |
+
"loss": 1.8872554779052735,
|
| 2131 |
+
"step": 3030
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 0.2643478260869565,
|
| 2135 |
+
"grad_norm": 0.04584391042590141,
|
| 2136 |
+
"learning_rate": 0.0001491305553157723,
|
| 2137 |
+
"loss": 1.883194351196289,
|
| 2138 |
+
"step": 3040
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 0.26521739130434785,
|
| 2142 |
+
"grad_norm": 0.04532180353999138,
|
| 2143 |
+
"learning_rate": 0.00014895763486737578,
|
| 2144 |
+
"loss": 1.8846633911132813,
|
| 2145 |
+
"step": 3050
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 0.26608695652173914,
|
| 2149 |
+
"grad_norm": 0.04704463481903076,
|
| 2150 |
+
"learning_rate": 0.000148784239391841,
|
| 2151 |
+
"loss": 1.8317132949829102,
|
| 2152 |
+
"step": 3060
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 0.2669565217391304,
|
| 2156 |
+
"grad_norm": 0.04611065611243248,
|
| 2157 |
+
"learning_rate": 0.000148610370229401,
|
| 2158 |
+
"loss": 1.9712543487548828,
|
| 2159 |
+
"step": 3070
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 0.2678260869565217,
|
| 2163 |
+
"grad_norm": 0.04814252257347107,
|
| 2164 |
+
"learning_rate": 0.00014843602872395025,
|
| 2165 |
+
"loss": 1.94564208984375,
|
| 2166 |
+
"step": 3080
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 0.26869565217391306,
|
| 2170 |
+
"grad_norm": 0.046260274946689606,
|
| 2171 |
+
"learning_rate": 0.00014826121622303398,
|
| 2172 |
+
"loss": 1.8456733703613282,
|
| 2173 |
+
"step": 3090
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 0.26956521739130435,
|
| 2177 |
+
"grad_norm": 0.04497453570365906,
|
| 2178 |
+
"learning_rate": 0.00014808593407783795,
|
| 2179 |
+
"loss": 1.8342208862304688,
|
| 2180 |
+
"step": 3100
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 0.27043478260869563,
|
| 2184 |
+
"grad_norm": 0.045834191143512726,
|
| 2185 |
+
"learning_rate": 0.00014791018364317807,
|
| 2186 |
+
"loss": 1.8588268280029296,
|
| 2187 |
+
"step": 3110
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 0.271304347826087,
|
| 2191 |
+
"grad_norm": 0.04720929265022278,
|
| 2192 |
+
"learning_rate": 0.00014773396627748963,
|
| 2193 |
+
"loss": 1.8708805084228515,
|
| 2194 |
+
"step": 3120
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 0.27217391304347827,
|
| 2198 |
+
"grad_norm": 0.044134072959423065,
|
| 2199 |
+
"learning_rate": 0.00014755728334281713,
|
| 2200 |
+
"loss": 1.8183509826660156,
|
| 2201 |
+
"step": 3130
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 0.27304347826086955,
|
| 2205 |
+
"grad_norm": 0.04427867382764816,
|
| 2206 |
+
"learning_rate": 0.00014738013620480355,
|
| 2207 |
+
"loss": 1.9084312438964843,
|
| 2208 |
+
"step": 3140
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 0.27391304347826084,
|
| 2212 |
+
"grad_norm": 0.04593170806765556,
|
| 2213 |
+
"learning_rate": 0.0001472025262326799,
|
| 2214 |
+
"loss": 1.9480342864990234,
|
| 2215 |
+
"step": 3150
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 0.2747826086956522,
|
| 2219 |
+
"grad_norm": 0.04596893861889839,
|
| 2220 |
+
"learning_rate": 0.0001470244547992546,
|
| 2221 |
+
"loss": 1.8195436477661133,
|
| 2222 |
+
"step": 3160
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 0.2756521739130435,
|
| 2226 |
+
"grad_norm": 0.04655667021870613,
|
| 2227 |
+
"learning_rate": 0.00014684592328090276,
|
| 2228 |
+
"loss": 1.9056392669677735,
|
| 2229 |
+
"step": 3170
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 0.27652173913043476,
|
| 2233 |
+
"grad_norm": 0.045723024755716324,
|
| 2234 |
+
"learning_rate": 0.00014666693305755575,
|
| 2235 |
+
"loss": 1.8669166564941406,
|
| 2236 |
+
"step": 3180
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 0.2773913043478261,
|
| 2240 |
+
"grad_norm": 0.04697049781680107,
|
| 2241 |
+
"learning_rate": 0.00014648748551269043,
|
| 2242 |
+
"loss": 1.8303703308105468,
|
| 2243 |
+
"step": 3190
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 0.2782608695652174,
|
| 2247 |
+
"grad_norm": 0.04601813852787018,
|
| 2248 |
+
"learning_rate": 0.00014630758203331836,
|
| 2249 |
+
"loss": 1.8695808410644532,
|
| 2250 |
+
"step": 3200
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 0.2791304347826087,
|
| 2254 |
+
"grad_norm": 0.04566989839076996,
|
| 2255 |
+
"learning_rate": 0.00014612722400997525,
|
| 2256 |
+
"loss": 1.8917108535766602,
|
| 2257 |
+
"step": 3210
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 0.28,
|
| 2261 |
+
"grad_norm": 0.04773734137415886,
|
| 2262 |
+
"learning_rate": 0.00014594641283671017,
|
| 2263 |
+
"loss": 1.895745086669922,
|
| 2264 |
+
"step": 3220
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 0.2808695652173913,
|
| 2268 |
+
"grad_norm": 0.04546461999416351,
|
| 2269 |
+
"learning_rate": 0.0001457651499110746,
|
| 2270 |
+
"loss": 1.8925960540771485,
|
| 2271 |
+
"step": 3230
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 0.2817391304347826,
|
| 2275 |
+
"grad_norm": 0.04565638676285744,
|
| 2276 |
+
"learning_rate": 0.00014558343663411197,
|
| 2277 |
+
"loss": 1.8597404479980468,
|
| 2278 |
+
"step": 3240
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 0.2826086956521739,
|
| 2282 |
+
"grad_norm": 0.045438554137945175,
|
| 2283 |
+
"learning_rate": 0.00014540127441034644,
|
| 2284 |
+
"loss": 1.8530876159667968,
|
| 2285 |
+
"step": 3250
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 0.28347826086956524,
|
| 2289 |
+
"grad_norm": 0.0473264716565609,
|
| 2290 |
+
"learning_rate": 0.00014521866464777233,
|
| 2291 |
+
"loss": 1.8691961288452148,
|
| 2292 |
+
"step": 3260
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 0.2843478260869565,
|
| 2296 |
+
"grad_norm": 0.04689843952655792,
|
| 2297 |
+
"learning_rate": 0.00014503560875784316,
|
| 2298 |
+
"loss": 1.8995010375976562,
|
| 2299 |
+
"step": 3270
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 0.2852173913043478,
|
| 2303 |
+
"grad_norm": 0.04655284807085991,
|
| 2304 |
+
"learning_rate": 0.00014485210815546066,
|
| 2305 |
+
"loss": 1.8623058319091796,
|
| 2306 |
+
"step": 3280
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 0.28608695652173916,
|
| 2310 |
+
"grad_norm": 0.04491620883345604,
|
| 2311 |
+
"learning_rate": 0.0001446681642589639,
|
| 2312 |
+
"loss": 1.912955093383789,
|
| 2313 |
+
"step": 3290
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 0.28695652173913044,
|
| 2317 |
+
"grad_norm": 0.04738801345229149,
|
| 2318 |
+
"learning_rate": 0.0001444837784901184,
|
| 2319 |
+
"loss": 1.8688699722290039,
|
| 2320 |
+
"step": 3300
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 0.28782608695652173,
|
| 2324 |
+
"grad_norm": 0.04594497010111809,
|
| 2325 |
+
"learning_rate": 0.00014429895227410493,
|
| 2326 |
+
"loss": 1.9313240051269531,
|
| 2327 |
+
"step": 3310
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 0.288695652173913,
|
| 2331 |
+
"grad_norm": 0.047535549849271774,
|
| 2332 |
+
"learning_rate": 0.0001441136870395088,
|
| 2333 |
+
"loss": 1.9088796615600585,
|
| 2334 |
+
"step": 3320
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 0.28956521739130436,
|
| 2338 |
+
"grad_norm": 0.04802543297410011,
|
| 2339 |
+
"learning_rate": 0.00014392798421830846,
|
| 2340 |
+
"loss": 1.8703914642333985,
|
| 2341 |
+
"step": 3330
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 0.29043478260869565,
|
| 2345 |
+
"grad_norm": 0.046663422137498856,
|
| 2346 |
+
"learning_rate": 0.00014374184524586477,
|
| 2347 |
+
"loss": 1.8450454711914062,
|
| 2348 |
+
"step": 3340
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 0.29130434782608694,
|
| 2352 |
+
"grad_norm": 0.046551477164030075,
|
| 2353 |
+
"learning_rate": 0.00014355527156090965,
|
| 2354 |
+
"loss": 1.9085540771484375,
|
| 2355 |
+
"step": 3350
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 0.2921739130434783,
|
| 2359 |
+
"grad_norm": 0.045060548931360245,
|
| 2360 |
+
"learning_rate": 0.00014336826460553513,
|
| 2361 |
+
"loss": 1.8219585418701172,
|
| 2362 |
+
"step": 3360
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 0.29304347826086957,
|
| 2366 |
+
"grad_norm": 0.04710334166884422,
|
| 2367 |
+
"learning_rate": 0.00014318082582518208,
|
| 2368 |
+
"loss": 1.9055330276489257,
|
| 2369 |
+
"step": 3370
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 0.29391304347826086,
|
| 2373 |
+
"grad_norm": 0.045783352106809616,
|
| 2374 |
+
"learning_rate": 0.00014299295666862912,
|
| 2375 |
+
"loss": 1.8600608825683593,
|
| 2376 |
+
"step": 3380
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 0.29478260869565215,
|
| 2380 |
+
"grad_norm": 0.04774748533964157,
|
| 2381 |
+
"learning_rate": 0.00014280465858798144,
|
| 2382 |
+
"loss": 1.8205097198486329,
|
| 2383 |
+
"step": 3390
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 0.2956521739130435,
|
| 2387 |
+
"grad_norm": 0.048090871423482895,
|
| 2388 |
+
"learning_rate": 0.00014261593303865943,
|
| 2389 |
+
"loss": 1.8969783782958984,
|
| 2390 |
+
"step": 3400
|
| 2391 |
+
}
|
| 2392 |
+
],
|
| 2393 |
+
"logging_steps": 10,
|
| 2394 |
+
"max_steps": 11500,
|
| 2395 |
+
"num_input_tokens_seen": 0,
|
| 2396 |
+
"num_train_epochs": 9223372036854775807,
|
| 2397 |
+
"save_steps": 100,
|
| 2398 |
+
"stateful_callbacks": {
|
| 2399 |
+
"TrainerControl": {
|
| 2400 |
+
"args": {
|
| 2401 |
+
"should_epoch_stop": false,
|
| 2402 |
+
"should_evaluate": false,
|
| 2403 |
+
"should_log": false,
|
| 2404 |
+
"should_save": true,
|
| 2405 |
+
"should_training_stop": false
|
| 2406 |
+
},
|
| 2407 |
+
"attributes": {}
|
| 2408 |
+
}
|
| 2409 |
+
},
|
| 2410 |
+
"total_flos": 3.019142697320448e+17,
|
| 2411 |
+
"train_batch_size": 8,
|
| 2412 |
+
"trial_name": null,
|
| 2413 |
+
"trial_params": null
|
| 2414 |
+
}
|