Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1953 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1953 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2735,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.018281535648994516,
|
| 13 |
+
"grad_norm": 16.712358474731445,
|
| 14 |
+
"learning_rate": 1.45985401459854e-05,
|
| 15 |
+
"loss": 1.4822,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.03656307129798903,
|
| 20 |
+
"grad_norm": 7.676208019256592,
|
| 21 |
+
"learning_rate": 2.91970802919708e-05,
|
| 22 |
+
"loss": 0.7397,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.054844606946983544,
|
| 27 |
+
"grad_norm": 2.2206971645355225,
|
| 28 |
+
"learning_rate": 4.379562043795621e-05,
|
| 29 |
+
"loss": 0.4701,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07312614259597806,
|
| 34 |
+
"grad_norm": 1.7638039588928223,
|
| 35 |
+
"learning_rate": 5.83941605839416e-05,
|
| 36 |
+
"loss": 0.2966,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.09140767824497258,
|
| 41 |
+
"grad_norm": 1.6052724123001099,
|
| 42 |
+
"learning_rate": 7.299270072992701e-05,
|
| 43 |
+
"loss": 0.2162,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.10968921389396709,
|
| 48 |
+
"grad_norm": 2.617760181427002,
|
| 49 |
+
"learning_rate": 8.759124087591242e-05,
|
| 50 |
+
"loss": 0.2019,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12797074954296161,
|
| 55 |
+
"grad_norm": 1.7860541343688965,
|
| 56 |
+
"learning_rate": 0.00010218978102189782,
|
| 57 |
+
"loss": 0.1457,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.14625228519195613,
|
| 62 |
+
"grad_norm": 1.4474908113479614,
|
| 63 |
+
"learning_rate": 0.0001167883211678832,
|
| 64 |
+
"loss": 0.1387,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.16453382084095064,
|
| 69 |
+
"grad_norm": 1.5035394430160522,
|
| 70 |
+
"learning_rate": 0.0001313868613138686,
|
| 71 |
+
"loss": 0.1363,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.18281535648994515,
|
| 76 |
+
"grad_norm": 1.4598884582519531,
|
| 77 |
+
"learning_rate": 0.00014598540145985403,
|
| 78 |
+
"loss": 0.1124,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.20109689213893966,
|
| 83 |
+
"grad_norm": 1.7308577299118042,
|
| 84 |
+
"learning_rate": 0.00016058394160583942,
|
| 85 |
+
"loss": 0.1215,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.21937842778793418,
|
| 90 |
+
"grad_norm": 1.5704491138458252,
|
| 91 |
+
"learning_rate": 0.00017518248175182484,
|
| 92 |
+
"loss": 0.1332,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.2376599634369287,
|
| 97 |
+
"grad_norm": 0.7519080638885498,
|
| 98 |
+
"learning_rate": 0.00018978102189781023,
|
| 99 |
+
"loss": 0.1015,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.25594149908592323,
|
| 104 |
+
"grad_norm": 1.0302314758300781,
|
| 105 |
+
"learning_rate": 0.00019999934198849153,
|
| 106 |
+
"loss": 0.1043,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.2742230347349177,
|
| 111 |
+
"grad_norm": 1.1439878940582275,
|
| 112 |
+
"learning_rate": 0.00019998764424701714,
|
| 113 |
+
"loss": 0.1105,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.29250457038391225,
|
| 118 |
+
"grad_norm": 0.8649179935455322,
|
| 119 |
+
"learning_rate": 0.00019996132599641746,
|
| 120 |
+
"loss": 0.0969,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.31078610603290674,
|
| 125 |
+
"grad_norm": 0.9194239377975464,
|
| 126 |
+
"learning_rate": 0.00019992039108503024,
|
| 127 |
+
"loss": 0.097,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.3290676416819013,
|
| 132 |
+
"grad_norm": 0.6259992718696594,
|
| 133 |
+
"learning_rate": 0.00019986484549848745,
|
| 134 |
+
"loss": 0.0853,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.3473491773308958,
|
| 139 |
+
"grad_norm": 1.0033239126205444,
|
| 140 |
+
"learning_rate": 0.00019979469735884026,
|
| 141 |
+
"loss": 0.0944,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.3656307129798903,
|
| 146 |
+
"grad_norm": 1.261385440826416,
|
| 147 |
+
"learning_rate": 0.00019970995692337114,
|
| 148 |
+
"loss": 0.1078,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.38391224862888484,
|
| 153 |
+
"grad_norm": 0.9231658577919006,
|
| 154 |
+
"learning_rate": 0.00019961063658309418,
|
| 155 |
+
"loss": 0.0821,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.40219378427787933,
|
| 160 |
+
"grad_norm": 0.996103048324585,
|
| 161 |
+
"learning_rate": 0.00019949675086094326,
|
| 162 |
+
"loss": 0.0911,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.42047531992687387,
|
| 167 |
+
"grad_norm": 0.9832742810249329,
|
| 168 |
+
"learning_rate": 0.0001993683164096483,
|
| 169 |
+
"loss": 0.0692,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.43875685557586835,
|
| 174 |
+
"grad_norm": 0.6472922563552856,
|
| 175 |
+
"learning_rate": 0.00019922535200930046,
|
| 176 |
+
"loss": 0.0706,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.4570383912248629,
|
| 181 |
+
"grad_norm": 0.5999054312705994,
|
| 182 |
+
"learning_rate": 0.00019906787856460581,
|
| 183 |
+
"loss": 0.0731,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.4753199268738574,
|
| 188 |
+
"grad_norm": 0.667738139629364,
|
| 189 |
+
"learning_rate": 0.00019889591910182876,
|
| 190 |
+
"loss": 0.0708,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.4936014625228519,
|
| 195 |
+
"grad_norm": 0.554964542388916,
|
| 196 |
+
"learning_rate": 0.0001987094987654251,
|
| 197 |
+
"loss": 0.0591,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.5118829981718465,
|
| 202 |
+
"grad_norm": 1.1600011587142944,
|
| 203 |
+
"learning_rate": 0.00019850864481436514,
|
| 204 |
+
"loss": 0.0795,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.5301645338208409,
|
| 209 |
+
"grad_norm": 0.6419970393180847,
|
| 210 |
+
"learning_rate": 0.00019829338661814797,
|
| 211 |
+
"loss": 0.0659,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5484460694698354,
|
| 216 |
+
"grad_norm": 0.735856831073761,
|
| 217 |
+
"learning_rate": 0.00019806375565250685,
|
| 218 |
+
"loss": 0.0724,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.56672760511883,
|
| 223 |
+
"grad_norm": 0.5395373106002808,
|
| 224 |
+
"learning_rate": 0.00019781978549480682,
|
| 225 |
+
"loss": 0.0626,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5850091407678245,
|
| 230 |
+
"grad_norm": 0.8947715759277344,
|
| 231 |
+
"learning_rate": 0.00019756151181913483,
|
| 232 |
+
"loss": 0.0601,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.603290676416819,
|
| 237 |
+
"grad_norm": 0.5075414180755615,
|
| 238 |
+
"learning_rate": 0.00019728897239108342,
|
| 239 |
+
"loss": 0.0691,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.6215722120658135,
|
| 244 |
+
"grad_norm": 1.3236219882965088,
|
| 245 |
+
"learning_rate": 0.00019700220706222858,
|
| 246 |
+
"loss": 0.0488,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.6398537477148081,
|
| 251 |
+
"grad_norm": 0.9153704047203064,
|
| 252 |
+
"learning_rate": 0.00019670125776430228,
|
| 253 |
+
"loss": 0.0622,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.6581352833638026,
|
| 258 |
+
"grad_norm": 0.6496918797492981,
|
| 259 |
+
"learning_rate": 0.00019638616850306133,
|
| 260 |
+
"loss": 0.0572,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.676416819012797,
|
| 265 |
+
"grad_norm": 0.6905117034912109,
|
| 266 |
+
"learning_rate": 0.00019605698535185266,
|
| 267 |
+
"loss": 0.0506,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.6946983546617916,
|
| 272 |
+
"grad_norm": 0.6502402424812317,
|
| 273 |
+
"learning_rate": 0.00019571375644487625,
|
| 274 |
+
"loss": 0.0528,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.7129798903107861,
|
| 279 |
+
"grad_norm": 0.7400691509246826,
|
| 280 |
+
"learning_rate": 0.0001953565319701469,
|
| 281 |
+
"loss": 0.0674,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.7312614259597806,
|
| 286 |
+
"grad_norm": 0.5896055698394775,
|
| 287 |
+
"learning_rate": 0.0001949853641621555,
|
| 288 |
+
"loss": 0.0471,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.7495429616087751,
|
| 293 |
+
"grad_norm": 0.4026470482349396,
|
| 294 |
+
"learning_rate": 0.00019460030729423114,
|
| 295 |
+
"loss": 0.0512,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.7678244972577697,
|
| 300 |
+
"grad_norm": 0.47957828640937805,
|
| 301 |
+
"learning_rate": 0.0001942014176706052,
|
| 302 |
+
"loss": 0.0629,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.7861060329067642,
|
| 307 |
+
"grad_norm": 0.4520862400531769,
|
| 308 |
+
"learning_rate": 0.00019378875361817817,
|
| 309 |
+
"loss": 0.0533,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.8043875685557587,
|
| 314 |
+
"grad_norm": 0.4732885956764221,
|
| 315 |
+
"learning_rate": 0.00019336237547799108,
|
| 316 |
+
"loss": 0.058,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.8226691042047533,
|
| 321 |
+
"grad_norm": 0.7703008651733398,
|
| 322 |
+
"learning_rate": 0.0001929223455964022,
|
| 323 |
+
"loss": 0.0532,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.8409506398537477,
|
| 328 |
+
"grad_norm": 0.45097994804382324,
|
| 329 |
+
"learning_rate": 0.00019246872831597055,
|
| 330 |
+
"loss": 0.0465,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.8592321755027422,
|
| 335 |
+
"grad_norm": 0.5736098289489746,
|
| 336 |
+
"learning_rate": 0.00019200158996604753,
|
| 337 |
+
"loss": 0.0487,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.8775137111517367,
|
| 342 |
+
"grad_norm": 0.7237376570701599,
|
| 343 |
+
"learning_rate": 0.0001915209988530779,
|
| 344 |
+
"loss": 0.0551,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.8957952468007313,
|
| 349 |
+
"grad_norm": 0.4645770192146301,
|
| 350 |
+
"learning_rate": 0.00019102702525061207,
|
| 351 |
+
"loss": 0.0495,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.9140767824497258,
|
| 356 |
+
"grad_norm": 0.5169672966003418,
|
| 357 |
+
"learning_rate": 0.00019051974138903027,
|
| 358 |
+
"loss": 0.0433,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.9323583180987203,
|
| 363 |
+
"grad_norm": 0.7457365989685059,
|
| 364 |
+
"learning_rate": 0.00018999922144498084,
|
| 365 |
+
"loss": 0.0518,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.9506398537477148,
|
| 370 |
+
"grad_norm": 0.5059699416160583,
|
| 371 |
+
"learning_rate": 0.00018946554153053395,
|
| 372 |
+
"loss": 0.0474,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.9689213893967094,
|
| 377 |
+
"grad_norm": 0.8174113035202026,
|
| 378 |
+
"learning_rate": 0.00018891877968205213,
|
| 379 |
+
"loss": 0.0517,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.9872029250457038,
|
| 384 |
+
"grad_norm": 0.5508332252502441,
|
| 385 |
+
"learning_rate": 0.00018835901584877973,
|
| 386 |
+
"loss": 0.0709,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.0054844606946984,
|
| 391 |
+
"grad_norm": 0.5709052681922913,
|
| 392 |
+
"learning_rate": 0.00018778633188115223,
|
| 393 |
+
"loss": 0.0484,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.023765996343693,
|
| 398 |
+
"grad_norm": 0.4354308247566223,
|
| 399 |
+
"learning_rate": 0.0001872008115188281,
|
| 400 |
+
"loss": 0.0544,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.0420475319926874,
|
| 405 |
+
"grad_norm": 0.535977303981781,
|
| 406 |
+
"learning_rate": 0.00018660254037844388,
|
| 407 |
+
"loss": 0.0562,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.0603290676416819,
|
| 412 |
+
"grad_norm": 0.2939574420452118,
|
| 413 |
+
"learning_rate": 0.00018599160594109522,
|
| 414 |
+
"loss": 0.0489,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.0786106032906764,
|
| 419 |
+
"grad_norm": 0.3677907884120941,
|
| 420 |
+
"learning_rate": 0.000185368097539545,
|
| 421 |
+
"loss": 0.0358,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.0968921389396709,
|
| 426 |
+
"grad_norm": 0.5382636785507202,
|
| 427 |
+
"learning_rate": 0.0001847321063451609,
|
| 428 |
+
"loss": 0.0395,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.1151736745886653,
|
| 433 |
+
"grad_norm": 0.457963764667511,
|
| 434 |
+
"learning_rate": 0.00018408372535458397,
|
| 435 |
+
"loss": 0.0523,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.13345521023766,
|
| 440 |
+
"grad_norm": 0.5560534000396729,
|
| 441 |
+
"learning_rate": 0.00018342304937613032,
|
| 442 |
+
"loss": 0.0531,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.1517367458866545,
|
| 447 |
+
"grad_norm": 0.6328279376029968,
|
| 448 |
+
"learning_rate": 0.00018275017501592818,
|
| 449 |
+
"loss": 0.0452,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.170018281535649,
|
| 454 |
+
"grad_norm": 0.45685553550720215,
|
| 455 |
+
"learning_rate": 0.0001820652006637915,
|
| 456 |
+
"loss": 0.0402,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.1882998171846435,
|
| 461 |
+
"grad_norm": 0.21566231548786163,
|
| 462 |
+
"learning_rate": 0.0001813682264788334,
|
| 463 |
+
"loss": 0.0401,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.206581352833638,
|
| 468 |
+
"grad_norm": 0.36770665645599365,
|
| 469 |
+
"learning_rate": 0.00018065935437482037,
|
| 470 |
+
"loss": 0.04,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.2248628884826325,
|
| 475 |
+
"grad_norm": 0.4096185863018036,
|
| 476 |
+
"learning_rate": 0.0001799386880052703,
|
| 477 |
+
"loss": 0.0352,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.2431444241316272,
|
| 482 |
+
"grad_norm": 0.4246453642845154,
|
| 483 |
+
"learning_rate": 0.00017920633274829575,
|
| 484 |
+
"loss": 0.045,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.2614259597806217,
|
| 489 |
+
"grad_norm": 0.4160013496875763,
|
| 490 |
+
"learning_rate": 0.00017846239569119528,
|
| 491 |
+
"loss": 0.0357,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.2797074954296161,
|
| 496 |
+
"grad_norm": 0.5409733653068542,
|
| 497 |
+
"learning_rate": 0.00017770698561479496,
|
| 498 |
+
"loss": 0.0376,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.2979890310786106,
|
| 503 |
+
"grad_norm": 0.22224466502666473,
|
| 504 |
+
"learning_rate": 0.00017694021297754188,
|
| 505 |
+
"loss": 0.041,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.3162705667276051,
|
| 510 |
+
"grad_norm": 0.5606803894042969,
|
| 511 |
+
"learning_rate": 0.00017616218989935272,
|
| 512 |
+
"loss": 0.0367,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.3345521023765996,
|
| 517 |
+
"grad_norm": 0.3131175935268402,
|
| 518 |
+
"learning_rate": 0.00017537303014521918,
|
| 519 |
+
"loss": 0.0466,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.352833638025594,
|
| 524 |
+
"grad_norm": 0.37444230914115906,
|
| 525 |
+
"learning_rate": 0.0001745728491085728,
|
| 526 |
+
"loss": 0.0401,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.3711151736745886,
|
| 531 |
+
"grad_norm": 0.6337727308273315,
|
| 532 |
+
"learning_rate": 0.0001737617637944119,
|
| 533 |
+
"loss": 0.0505,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.389396709323583,
|
| 538 |
+
"grad_norm": 0.5669440627098083,
|
| 539 |
+
"learning_rate": 0.00017293989280219274,
|
| 540 |
+
"loss": 0.0372,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.4076782449725778,
|
| 545 |
+
"grad_norm": 0.388346791267395,
|
| 546 |
+
"learning_rate": 0.00017210735630848745,
|
| 547 |
+
"loss": 0.035,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.4259597806215722,
|
| 552 |
+
"grad_norm": 0.5280373096466064,
|
| 553 |
+
"learning_rate": 0.00017126427604941148,
|
| 554 |
+
"loss": 0.0466,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.4442413162705667,
|
| 559 |
+
"grad_norm": 0.565298855304718,
|
| 560 |
+
"learning_rate": 0.00017041077530282294,
|
| 561 |
+
"loss": 0.0365,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.4625228519195612,
|
| 566 |
+
"grad_norm": 0.35680803656578064,
|
| 567 |
+
"learning_rate": 0.00016954697887029655,
|
| 568 |
+
"loss": 0.0383,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.4808043875685557,
|
| 573 |
+
"grad_norm": 0.42788997292518616,
|
| 574 |
+
"learning_rate": 0.00016867301305887474,
|
| 575 |
+
"loss": 0.0337,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.4990859232175504,
|
| 580 |
+
"grad_norm": 0.43233945965766907,
|
| 581 |
+
"learning_rate": 0.00016778900566259865,
|
| 582 |
+
"loss": 0.0505,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.517367458866545,
|
| 587 |
+
"grad_norm": 0.4589940905570984,
|
| 588 |
+
"learning_rate": 0.0001668950859438216,
|
| 589 |
+
"loss": 0.0438,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.5356489945155394,
|
| 594 |
+
"grad_norm": 0.48594310879707336,
|
| 595 |
+
"learning_rate": 0.00016599138461430814,
|
| 596 |
+
"loss": 0.0323,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.5539305301645339,
|
| 601 |
+
"grad_norm": 0.31333279609680176,
|
| 602 |
+
"learning_rate": 0.00016507803381612076,
|
| 603 |
+
"loss": 0.0393,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.5722120658135283,
|
| 608 |
+
"grad_norm": 0.49847719073295593,
|
| 609 |
+
"learning_rate": 0.00016415516710229766,
|
| 610 |
+
"loss": 0.0453,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.5904936014625228,
|
| 615 |
+
"grad_norm": 0.4276566505432129,
|
| 616 |
+
"learning_rate": 0.00016322291941732442,
|
| 617 |
+
"loss": 0.0362,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.6087751371115173,
|
| 622 |
+
"grad_norm": 0.47734275460243225,
|
| 623 |
+
"learning_rate": 0.0001622814270774018,
|
| 624 |
+
"loss": 0.0349,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.6270566727605118,
|
| 629 |
+
"grad_norm": 0.24307364225387573,
|
| 630 |
+
"learning_rate": 0.00016133082775051313,
|
| 631 |
+
"loss": 0.0365,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.6453382084095063,
|
| 636 |
+
"grad_norm": 0.4327755272388458,
|
| 637 |
+
"learning_rate": 0.00016037126043629422,
|
| 638 |
+
"loss": 0.0318,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.6636197440585008,
|
| 643 |
+
"grad_norm": 0.2253831923007965,
|
| 644 |
+
"learning_rate": 0.0001594028654457083,
|
| 645 |
+
"loss": 0.0324,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.6819012797074955,
|
| 650 |
+
"grad_norm": 0.42007511854171753,
|
| 651 |
+
"learning_rate": 0.0001584257843805293,
|
| 652 |
+
"loss": 0.0387,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.70018281535649,
|
| 657 |
+
"grad_norm": 0.5654010772705078,
|
| 658 |
+
"learning_rate": 0.00015744016011263638,
|
| 659 |
+
"loss": 0.0461,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.7184643510054844,
|
| 664 |
+
"grad_norm": 0.5979740619659424,
|
| 665 |
+
"learning_rate": 0.00015644613676312288,
|
| 666 |
+
"loss": 0.0288,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.736745886654479,
|
| 671 |
+
"grad_norm": 0.6250779628753662,
|
| 672 |
+
"learning_rate": 0.00015544385968122227,
|
| 673 |
+
"loss": 0.0339,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.7550274223034736,
|
| 678 |
+
"grad_norm": 0.4420310854911804,
|
| 679 |
+
"learning_rate": 0.00015443347542305484,
|
| 680 |
+
"loss": 0.0446,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.7733089579524681,
|
| 685 |
+
"grad_norm": 0.4242953956127167,
|
| 686 |
+
"learning_rate": 0.0001534151317301979,
|
| 687 |
+
"loss": 0.0402,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.7915904936014626,
|
| 692 |
+
"grad_norm": 0.2853521406650543,
|
| 693 |
+
"learning_rate": 0.00015238897750808242,
|
| 694 |
+
"loss": 0.0367,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.809872029250457,
|
| 699 |
+
"grad_norm": 0.5415486693382263,
|
| 700 |
+
"learning_rate": 0.00015135516280421945,
|
| 701 |
+
"loss": 0.0312,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.8281535648994516,
|
| 706 |
+
"grad_norm": 0.3944428265094757,
|
| 707 |
+
"learning_rate": 0.00015031383878626016,
|
| 708 |
+
"loss": 0.0293,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.846435100548446,
|
| 713 |
+
"grad_norm": 0.42964455485343933,
|
| 714 |
+
"learning_rate": 0.00014926515771989104,
|
| 715 |
+
"loss": 0.0462,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.8647166361974405,
|
| 720 |
+
"grad_norm": 0.3574308454990387,
|
| 721 |
+
"learning_rate": 0.00014820927294656973,
|
| 722 |
+
"loss": 0.0358,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.882998171846435,
|
| 727 |
+
"grad_norm": 0.38193315267562866,
|
| 728 |
+
"learning_rate": 0.00014714633886110242,
|
| 729 |
+
"loss": 0.0393,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.9012797074954295,
|
| 734 |
+
"grad_norm": 0.4956030249595642,
|
| 735 |
+
"learning_rate": 0.00014607651088906809,
|
| 736 |
+
"loss": 0.0312,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.919561243144424,
|
| 741 |
+
"grad_norm": 0.4244064688682556,
|
| 742 |
+
"learning_rate": 0.00014499994546409152,
|
| 743 |
+
"loss": 0.031,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.9378427787934185,
|
| 748 |
+
"grad_norm": 0.46385011076927185,
|
| 749 |
+
"learning_rate": 0.00014391680000496932,
|
| 750 |
+
"loss": 0.0424,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.9561243144424132,
|
| 755 |
+
"grad_norm": 0.5440361499786377,
|
| 756 |
+
"learning_rate": 0.0001428272328926512,
|
| 757 |
+
"loss": 0.0328,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.9744058500914077,
|
| 762 |
+
"grad_norm": 0.3221015930175781,
|
| 763 |
+
"learning_rate": 0.00014173140344708152,
|
| 764 |
+
"loss": 0.0424,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.9926873857404022,
|
| 769 |
+
"grad_norm": 0.520367443561554,
|
| 770 |
+
"learning_rate": 0.00014062947190390262,
|
| 771 |
+
"loss": 0.0396,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 2.010968921389397,
|
| 776 |
+
"grad_norm": 0.29480573534965515,
|
| 777 |
+
"learning_rate": 0.0001395215993910249,
|
| 778 |
+
"loss": 0.0351,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 2.0292504570383914,
|
| 783 |
+
"grad_norm": 0.35179761052131653,
|
| 784 |
+
"learning_rate": 0.00013840794790506616,
|
| 785 |
+
"loss": 0.0271,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 2.047531992687386,
|
| 790 |
+
"grad_norm": 0.377270370721817,
|
| 791 |
+
"learning_rate": 0.00013728868028766377,
|
| 792 |
+
"loss": 0.0311,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 2.0658135283363803,
|
| 797 |
+
"grad_norm": 0.4772701859474182,
|
| 798 |
+
"learning_rate": 0.0001361639602016637,
|
| 799 |
+
"loss": 0.0372,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.084095063985375,
|
| 804 |
+
"grad_norm": 0.30298906564712524,
|
| 805 |
+
"learning_rate": 0.000135033952107189,
|
| 806 |
+
"loss": 0.0255,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 2.1023765996343693,
|
| 811 |
+
"grad_norm": 0.39370113611221313,
|
| 812 |
+
"learning_rate": 0.00013389882123759206,
|
| 813 |
+
"loss": 0.0327,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 2.1206581352833638,
|
| 818 |
+
"grad_norm": 0.2912181317806244,
|
| 819 |
+
"learning_rate": 0.00013275873357529368,
|
| 820 |
+
"loss": 0.0268,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 2.1389396709323583,
|
| 825 |
+
"grad_norm": 0.29357820749282837,
|
| 826 |
+
"learning_rate": 0.00013161385582751247,
|
| 827 |
+
"loss": 0.0273,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 2.1572212065813527,
|
| 832 |
+
"grad_norm": 0.3242945075035095,
|
| 833 |
+
"learning_rate": 0.00013046435540188848,
|
| 834 |
+
"loss": 0.0296,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.1755027422303472,
|
| 839 |
+
"grad_norm": 1.168150544166565,
|
| 840 |
+
"learning_rate": 0.00012931040038200435,
|
| 841 |
+
"loss": 0.0416,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.1937842778793417,
|
| 846 |
+
"grad_norm": 0.3501128852367401,
|
| 847 |
+
"learning_rate": 0.00012815215950280753,
|
| 848 |
+
"loss": 0.0379,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 2.212065813528336,
|
| 853 |
+
"grad_norm": 0.46127256751060486,
|
| 854 |
+
"learning_rate": 0.0001269898021259373,
|
| 855 |
+
"loss": 0.0372,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 2.2303473491773307,
|
| 860 |
+
"grad_norm": 0.4480052888393402,
|
| 861 |
+
"learning_rate": 0.0001258234982149604,
|
| 862 |
+
"loss": 0.0366,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 2.2486288848263256,
|
| 867 |
+
"grad_norm": 0.38535383343696594,
|
| 868 |
+
"learning_rate": 0.0001246534183105181,
|
| 869 |
+
"loss": 0.0289,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 2.26691042047532,
|
| 874 |
+
"grad_norm": 0.39918404817581177,
|
| 875 |
+
"learning_rate": 0.00012347973350538936,
|
| 876 |
+
"loss": 0.029,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 2.2851919561243146,
|
| 881 |
+
"grad_norm": 0.27928563952445984,
|
| 882 |
+
"learning_rate": 0.00012230261541947316,
|
| 883 |
+
"loss": 0.0262,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.303473491773309,
|
| 888 |
+
"grad_norm": 0.43867453932762146,
|
| 889 |
+
"learning_rate": 0.00012112223617469372,
|
| 890 |
+
"loss": 0.0227,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 2.3217550274223036,
|
| 895 |
+
"grad_norm": 0.3848976194858551,
|
| 896 |
+
"learning_rate": 0.00011993876836983198,
|
| 897 |
+
"loss": 0.0251,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 2.340036563071298,
|
| 902 |
+
"grad_norm": 0.3365519046783447,
|
| 903 |
+
"learning_rate": 0.0001187523850552881,
|
| 904 |
+
"loss": 0.0345,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 2.3583180987202925,
|
| 909 |
+
"grad_norm": 0.3406737446784973,
|
| 910 |
+
"learning_rate": 0.00011756325970777717,
|
| 911 |
+
"loss": 0.0273,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.376599634369287,
|
| 916 |
+
"grad_norm": 0.28142690658569336,
|
| 917 |
+
"learning_rate": 0.00011637156620496308,
|
| 918 |
+
"loss": 0.0275,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 2.3948811700182815,
|
| 923 |
+
"grad_norm": 0.36976391077041626,
|
| 924 |
+
"learning_rate": 0.00011517747880003335,
|
| 925 |
+
"loss": 0.0243,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.413162705667276,
|
| 930 |
+
"grad_norm": 0.22825664281845093,
|
| 931 |
+
"learning_rate": 0.00011398117209621966,
|
| 932 |
+
"loss": 0.0278,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 2.4314442413162705,
|
| 937 |
+
"grad_norm": 0.3394540548324585,
|
| 938 |
+
"learning_rate": 0.00011278282102126633,
|
| 939 |
+
"loss": 0.0357,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 2.449725776965265,
|
| 944 |
+
"grad_norm": 0.26682886481285095,
|
| 945 |
+
"learning_rate": 0.00011158260080185226,
|
| 946 |
+
"loss": 0.0407,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 2.4680073126142594,
|
| 951 |
+
"grad_norm": 0.23459388315677643,
|
| 952 |
+
"learning_rate": 0.00011038068693796846,
|
| 953 |
+
"loss": 0.0263,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 2.4862888482632544,
|
| 958 |
+
"grad_norm": 0.32797005772590637,
|
| 959 |
+
"learning_rate": 0.00010917725517725608,
|
| 960 |
+
"loss": 0.0354,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 2.504570383912249,
|
| 965 |
+
"grad_norm": 0.2672847509384155,
|
| 966 |
+
"learning_rate": 0.00010797248148930783,
|
| 967 |
+
"loss": 0.0203,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.5228519195612433,
|
| 972 |
+
"grad_norm": 0.34542036056518555,
|
| 973 |
+
"learning_rate": 0.00010676654203993732,
|
| 974 |
+
"loss": 0.0246,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 2.541133455210238,
|
| 979 |
+
"grad_norm": 0.5064176321029663,
|
| 980 |
+
"learning_rate": 0.00010555961316541946,
|
| 981 |
+
"loss": 0.0276,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 2.5594149908592323,
|
| 986 |
+
"grad_norm": 0.34542617201805115,
|
| 987 |
+
"learning_rate": 0.00010435187134670607,
|
| 988 |
+
"loss": 0.0238,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.577696526508227,
|
| 993 |
+
"grad_norm": 0.3336438238620758,
|
| 994 |
+
"learning_rate": 0.00010314349318362015,
|
| 995 |
+
"loss": 0.0353,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 2.5959780621572213,
|
| 1000 |
+
"grad_norm": 0.22887752950191498,
|
| 1001 |
+
"learning_rate": 0.00010193465536903307,
|
| 1002 |
+
"loss": 0.028,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 2.6142595978062158,
|
| 1007 |
+
"grad_norm": 0.1399448662996292,
|
| 1008 |
+
"learning_rate": 0.00010072553466302784,
|
| 1009 |
+
"loss": 0.028,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.6325411334552102,
|
| 1014 |
+
"grad_norm": 0.36335644125938416,
|
| 1015 |
+
"learning_rate": 9.951630786705279e-05,
|
| 1016 |
+
"loss": 0.0196,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 2.6508226691042047,
|
| 1021 |
+
"grad_norm": 0.22947153449058533,
|
| 1022 |
+
"learning_rate": 9.830715179806905e-05,
|
| 1023 |
+
"loss": 0.0275,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 2.669104204753199,
|
| 1028 |
+
"grad_norm": 0.21563003957271576,
|
| 1029 |
+
"learning_rate": 9.709824326269576e-05,
|
| 1030 |
+
"loss": 0.0216,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 2.6873857404021937,
|
| 1035 |
+
"grad_norm": 0.3260309100151062,
|
| 1036 |
+
"learning_rate": 9.5889759031357e-05,
|
| 1037 |
+
"loss": 0.018,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 2.705667276051188,
|
| 1042 |
+
"grad_norm": 0.15418443083763123,
|
| 1043 |
+
"learning_rate": 9.468187581243378e-05,
|
| 1044 |
+
"loss": 0.0244,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 2.7239488117001827,
|
| 1049 |
+
"grad_norm": 0.2873231768608093,
|
| 1050 |
+
"learning_rate": 9.347477022642503e-05,
|
| 1051 |
+
"loss": 0.0186,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.742230347349177,
|
| 1056 |
+
"grad_norm": 0.2715139091014862,
|
| 1057 |
+
"learning_rate": 9.226861878012197e-05,
|
| 1058 |
+
"loss": 0.0273,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 2.7605118829981716,
|
| 1063 |
+
"grad_norm": 0.17074620723724365,
|
| 1064 |
+
"learning_rate": 9.106359784079832e-05,
|
| 1065 |
+
"loss": 0.0174,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 2.778793418647166,
|
| 1070 |
+
"grad_norm": 0.2897492051124573,
|
| 1071 |
+
"learning_rate": 8.985988361042153e-05,
|
| 1072 |
+
"loss": 0.0283,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 2.797074954296161,
|
| 1077 |
+
"grad_norm": 0.5155644416809082,
|
| 1078 |
+
"learning_rate": 8.8657652099888e-05,
|
| 1079 |
+
"loss": 0.0216,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 2.8153564899451555,
|
| 1084 |
+
"grad_norm": 0.33276352286338806,
|
| 1085 |
+
"learning_rate": 8.745707910328615e-05,
|
| 1086 |
+
"loss": 0.0245,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 2.83363802559415,
|
| 1091 |
+
"grad_norm": 0.4756206274032593,
|
| 1092 |
+
"learning_rate": 8.625834017219113e-05,
|
| 1093 |
+
"loss": 0.0303,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.8519195612431445,
|
| 1098 |
+
"grad_norm": 0.2755451202392578,
|
| 1099 |
+
"learning_rate": 8.506161058999541e-05,
|
| 1100 |
+
"loss": 0.0199,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 2.870201096892139,
|
| 1105 |
+
"grad_norm": 0.26369351148605347,
|
| 1106 |
+
"learning_rate": 8.386706534627805e-05,
|
| 1107 |
+
"loss": 0.0204,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 2.8884826325411335,
|
| 1112 |
+
"grad_norm": 0.2358650118112564,
|
| 1113 |
+
"learning_rate": 8.267487911121715e-05,
|
| 1114 |
+
"loss": 0.0211,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 2.906764168190128,
|
| 1119 |
+
"grad_norm": 0.22182169556617737,
|
| 1120 |
+
"learning_rate": 8.148522621004926e-05,
|
| 1121 |
+
"loss": 0.0233,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 2.9250457038391224,
|
| 1126 |
+
"grad_norm": 0.30960527062416077,
|
| 1127 |
+
"learning_rate": 8.029828059757875e-05,
|
| 1128 |
+
"loss": 0.0243,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 2.943327239488117,
|
| 1133 |
+
"grad_norm": 0.38207757472991943,
|
| 1134 |
+
"learning_rate": 7.91142158327417e-05,
|
| 1135 |
+
"loss": 0.0295,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.9616087751371114,
|
| 1140 |
+
"grad_norm": 0.24521781504154205,
|
| 1141 |
+
"learning_rate": 7.793320505322761e-05,
|
| 1142 |
+
"loss": 0.0206,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 2.979890310786106,
|
| 1147 |
+
"grad_norm": 0.3253994286060333,
|
| 1148 |
+
"learning_rate": 7.675542095016256e-05,
|
| 1149 |
+
"loss": 0.026,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 2.998171846435101,
|
| 1154 |
+
"grad_norm": 0.3253840208053589,
|
| 1155 |
+
"learning_rate": 7.558103574285779e-05,
|
| 1156 |
+
"loss": 0.0219,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 3.016453382084095,
|
| 1161 |
+
"grad_norm": 0.2342890352010727,
|
| 1162 |
+
"learning_rate": 7.441022115362729e-05,
|
| 1163 |
+
"loss": 0.0181,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 3.03473491773309,
|
| 1168 |
+
"grad_norm": 0.2249564677476883,
|
| 1169 |
+
"learning_rate": 7.324314838267796e-05,
|
| 1170 |
+
"loss": 0.0228,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 3.0530164533820843,
|
| 1175 |
+
"grad_norm": 0.24722999334335327,
|
| 1176 |
+
"learning_rate": 7.207998808307628e-05,
|
| 1177 |
+
"loss": 0.018,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 3.0712979890310788,
|
| 1182 |
+
"grad_norm": 0.22779327630996704,
|
| 1183 |
+
"learning_rate": 7.092091033579475e-05,
|
| 1184 |
+
"loss": 0.0193,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 3.0895795246800732,
|
| 1189 |
+
"grad_norm": 0.34452179074287415,
|
| 1190 |
+
"learning_rate": 6.976608462484226e-05,
|
| 1191 |
+
"loss": 0.0327,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 3.1078610603290677,
|
| 1196 |
+
"grad_norm": 0.30508124828338623,
|
| 1197 |
+
"learning_rate": 6.861567981248142e-05,
|
| 1198 |
+
"loss": 0.0261,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 3.126142595978062,
|
| 1203 |
+
"grad_norm": 0.319670706987381,
|
| 1204 |
+
"learning_rate": 6.746986411453717e-05,
|
| 1205 |
+
"loss": 0.0189,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 3.1444241316270567,
|
| 1210 |
+
"grad_norm": 0.35580283403396606,
|
| 1211 |
+
"learning_rate": 6.632880507579957e-05,
|
| 1212 |
+
"loss": 0.0242,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 3.162705667276051,
|
| 1217 |
+
"grad_norm": 0.3020285964012146,
|
| 1218 |
+
"learning_rate": 6.519266954552502e-05,
|
| 1219 |
+
"loss": 0.0176,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 3.1809872029250457,
|
| 1224 |
+
"grad_norm": 0.27105554938316345,
|
| 1225 |
+
"learning_rate": 6.406162365303882e-05,
|
| 1226 |
+
"loss": 0.0268,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 3.19926873857404,
|
| 1231 |
+
"grad_norm": 0.20928241312503815,
|
| 1232 |
+
"learning_rate": 6.293583278344361e-05,
|
| 1233 |
+
"loss": 0.0206,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 3.2175502742230346,
|
| 1238 |
+
"grad_norm": 0.2314785271883011,
|
| 1239 |
+
"learning_rate": 6.181546155343579e-05,
|
| 1240 |
+
"loss": 0.0198,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 3.235831809872029,
|
| 1245 |
+
"grad_norm": 0.2732461988925934,
|
| 1246 |
+
"learning_rate": 6.070067378723501e-05,
|
| 1247 |
+
"loss": 0.0177,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 3.2541133455210236,
|
| 1252 |
+
"grad_norm": 0.17697979509830475,
|
| 1253 |
+
"learning_rate": 5.959163249262913e-05,
|
| 1254 |
+
"loss": 0.0155,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 3.272394881170018,
|
| 1259 |
+
"grad_norm": 0.24429567158222198,
|
| 1260 |
+
"learning_rate": 5.848849983713894e-05,
|
| 1261 |
+
"loss": 0.0212,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 3.2906764168190126,
|
| 1266 |
+
"grad_norm": 0.36660096049308777,
|
| 1267 |
+
"learning_rate": 5.739143712430521e-05,
|
| 1268 |
+
"loss": 0.0281,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 3.3089579524680075,
|
| 1273 |
+
"grad_norm": 0.2895634174346924,
|
| 1274 |
+
"learning_rate": 5.630060477010253e-05,
|
| 1275 |
+
"loss": 0.018,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 3.327239488117002,
|
| 1280 |
+
"grad_norm": 0.3412606418132782,
|
| 1281 |
+
"learning_rate": 5.5216162279482964e-05,
|
| 1282 |
+
"loss": 0.0134,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 3.3455210237659965,
|
| 1287 |
+
"grad_norm": 0.22716091573238373,
|
| 1288 |
+
"learning_rate": 5.4138268223052326e-05,
|
| 1289 |
+
"loss": 0.016,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 3.363802559414991,
|
| 1294 |
+
"grad_norm": 0.24945920705795288,
|
| 1295 |
+
"learning_rate": 5.306708021388378e-05,
|
| 1296 |
+
"loss": 0.0208,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 3.3820840950639854,
|
| 1301 |
+
"grad_norm": 0.24487105011940002,
|
| 1302 |
+
"learning_rate": 5.200275488447104e-05,
|
| 1303 |
+
"loss": 0.018,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 3.40036563071298,
|
| 1308 |
+
"grad_norm": 0.24816852807998657,
|
| 1309 |
+
"learning_rate": 5.094544786382522e-05,
|
| 1310 |
+
"loss": 0.0159,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 3.4186471663619744,
|
| 1315 |
+
"grad_norm": 0.1848219782114029,
|
| 1316 |
+
"learning_rate": 4.989531375471805e-05,
|
| 1317 |
+
"loss": 0.0142,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 3.436928702010969,
|
| 1322 |
+
"grad_norm": 0.19923894107341766,
|
| 1323 |
+
"learning_rate": 4.885250611107558e-05,
|
| 1324 |
+
"loss": 0.0214,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 3.4552102376599634,
|
| 1329 |
+
"grad_norm": 0.1752861738204956,
|
| 1330 |
+
"learning_rate": 4.7817177415524796e-05,
|
| 1331 |
+
"loss": 0.0198,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 3.473491773308958,
|
| 1336 |
+
"grad_norm": 0.3053307831287384,
|
| 1337 |
+
"learning_rate": 4.678947905709744e-05,
|
| 1338 |
+
"loss": 0.0225,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 3.4917733089579523,
|
| 1343 |
+
"grad_norm": 0.19800381362438202,
|
| 1344 |
+
"learning_rate": 4.576956130909317e-05,
|
| 1345 |
+
"loss": 0.016,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 3.510054844606947,
|
| 1350 |
+
"grad_norm": 0.1873503029346466,
|
| 1351 |
+
"learning_rate": 4.475757330710621e-05,
|
| 1352 |
+
"loss": 0.0144,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 3.5283363802559418,
|
| 1357 |
+
"grad_norm": 0.23367895185947418,
|
| 1358 |
+
"learning_rate": 4.375366302721825e-05,
|
| 1359 |
+
"loss": 0.0161,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 3.5466179159049362,
|
| 1364 |
+
"grad_norm": 0.17103944718837738,
|
| 1365 |
+
"learning_rate": 4.2757977264361046e-05,
|
| 1366 |
+
"loss": 0.0146,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 3.5648994515539307,
|
| 1371 |
+
"grad_norm": 0.2473006546497345,
|
| 1372 |
+
"learning_rate": 4.177066161085148e-05,
|
| 1373 |
+
"loss": 0.0184,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 3.583180987202925,
|
| 1378 |
+
"grad_norm": 0.31398236751556396,
|
| 1379 |
+
"learning_rate": 4.0791860435102524e-05,
|
| 1380 |
+
"loss": 0.0146,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 3.6014625228519197,
|
| 1385 |
+
"grad_norm": 0.33835136890411377,
|
| 1386 |
+
"learning_rate": 3.982171686051334e-05,
|
| 1387 |
+
"loss": 0.021,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 3.619744058500914,
|
| 1392 |
+
"grad_norm": 0.16258537769317627,
|
| 1393 |
+
"learning_rate": 3.8860372744541407e-05,
|
| 1394 |
+
"loss": 0.0196,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 3.6380255941499087,
|
| 1399 |
+
"grad_norm": 0.3083174228668213,
|
| 1400 |
+
"learning_rate": 3.790796865795947e-05,
|
| 1401 |
+
"loss": 0.0152,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 3.656307129798903,
|
| 1406 |
+
"grad_norm": 0.21282333135604858,
|
| 1407 |
+
"learning_rate": 3.696464386430093e-05,
|
| 1408 |
+
"loss": 0.0215,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 3.6745886654478976,
|
| 1413 |
+
"grad_norm": 0.20800185203552246,
|
| 1414 |
+
"learning_rate": 3.6030536299496395e-05,
|
| 1415 |
+
"loss": 0.0155,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 3.692870201096892,
|
| 1420 |
+
"grad_norm": 0.251663476228714,
|
| 1421 |
+
"learning_rate": 3.5105782551704145e-05,
|
| 1422 |
+
"loss": 0.0222,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 3.7111517367458866,
|
| 1427 |
+
"grad_norm": 0.24097998440265656,
|
| 1428 |
+
"learning_rate": 3.419051784133773e-05,
|
| 1429 |
+
"loss": 0.0142,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 3.729433272394881,
|
| 1434 |
+
"grad_norm": 0.18417520821094513,
|
| 1435 |
+
"learning_rate": 3.328487600129371e-05,
|
| 1436 |
+
"loss": 0.0147,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 3.7477148080438756,
|
| 1441 |
+
"grad_norm": 0.18106205761432648,
|
| 1442 |
+
"learning_rate": 3.2388989457382126e-05,
|
| 1443 |
+
"loss": 0.0125,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 3.76599634369287,
|
| 1448 |
+
"grad_norm": 0.14622414112091064,
|
| 1449 |
+
"learning_rate": 3.1502989208962855e-05,
|
| 1450 |
+
"loss": 0.0151,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 3.7842778793418645,
|
| 1455 |
+
"grad_norm": 0.29628556966781616,
|
| 1456 |
+
"learning_rate": 3.062700480979046e-05,
|
| 1457 |
+
"loss": 0.0206,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 3.802559414990859,
|
| 1462 |
+
"grad_norm": 0.26881730556488037,
|
| 1463 |
+
"learning_rate": 2.9761164349070315e-05,
|
| 1464 |
+
"loss": 0.0176,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 3.8208409506398535,
|
| 1469 |
+
"grad_norm": 0.4180646240711212,
|
| 1470 |
+
"learning_rate": 2.8905594432729055e-05,
|
| 1471 |
+
"loss": 0.0179,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 3.839122486288848,
|
| 1476 |
+
"grad_norm": 0.25500163435935974,
|
| 1477 |
+
"learning_rate": 2.8060420164902012e-05,
|
| 1478 |
+
"loss": 0.0142,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 3.857404021937843,
|
| 1483 |
+
"grad_norm": 0.21968974173069,
|
| 1484 |
+
"learning_rate": 2.7225765129639836e-05,
|
| 1485 |
+
"loss": 0.0161,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 3.8756855575868374,
|
| 1490 |
+
"grad_norm": 0.24668078124523163,
|
| 1491 |
+
"learning_rate": 2.6401751372837813e-05,
|
| 1492 |
+
"loss": 0.0217,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 3.893967093235832,
|
| 1497 |
+
"grad_norm": 0.2258848249912262,
|
| 1498 |
+
"learning_rate": 2.5588499384389865e-05,
|
| 1499 |
+
"loss": 0.0178,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 3.9122486288848264,
|
| 1504 |
+
"grad_norm": 0.1784961074590683,
|
| 1505 |
+
"learning_rate": 2.478612808057018e-05,
|
| 1506 |
+
"loss": 0.0114,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 3.930530164533821,
|
| 1511 |
+
"grad_norm": 0.28832298517227173,
|
| 1512 |
+
"learning_rate": 2.3994754786644923e-05,
|
| 1513 |
+
"loss": 0.0109,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 3.9488117001828154,
|
| 1518 |
+
"grad_norm": 0.12029292434453964,
|
| 1519 |
+
"learning_rate": 2.3214495219716436e-05,
|
| 1520 |
+
"loss": 0.0211,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 3.96709323583181,
|
| 1525 |
+
"grad_norm": 0.19231897592544556,
|
| 1526 |
+
"learning_rate": 2.2445463471802785e-05,
|
| 1527 |
+
"loss": 0.0098,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 3.9853747714808043,
|
| 1532 |
+
"grad_norm": 0.08982887864112854,
|
| 1533 |
+
"learning_rate": 2.1687771993155004e-05,
|
| 1534 |
+
"loss": 0.0077,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 4.003656307129799,
|
| 1539 |
+
"grad_norm": 0.21466206014156342,
|
| 1540 |
+
"learning_rate": 2.0941531575813988e-05,
|
| 1541 |
+
"loss": 0.0159,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 4.021937842778794,
|
| 1546 |
+
"grad_norm": 0.17917244136333466,
|
| 1547 |
+
"learning_rate": 2.0206851337410415e-05,
|
| 1548 |
+
"loss": 0.0139,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 4.040219378427788,
|
| 1553 |
+
"grad_norm": 0.08883915841579437,
|
| 1554 |
+
"learning_rate": 1.9483838705209012e-05,
|
| 1555 |
+
"loss": 0.0152,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 4.058500914076783,
|
| 1560 |
+
"grad_norm": 0.16674405336380005,
|
| 1561 |
+
"learning_rate": 1.8772599400400258e-05,
|
| 1562 |
+
"loss": 0.0196,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 4.076782449725777,
|
| 1567 |
+
"grad_norm": 0.10342701524496078,
|
| 1568 |
+
"learning_rate": 1.807323742264162e-05,
|
| 1569 |
+
"loss": 0.0161,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 4.095063985374772,
|
| 1574 |
+
"grad_norm": 0.1896440088748932,
|
| 1575 |
+
"learning_rate": 1.7385855034850184e-05,
|
| 1576 |
+
"loss": 0.0122,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 4.113345521023766,
|
| 1581 |
+
"grad_norm": 0.16498374938964844,
|
| 1582 |
+
"learning_rate": 1.6710552748249598e-05,
|
| 1583 |
+
"loss": 0.0133,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 4.131627056672761,
|
| 1588 |
+
"grad_norm": 0.17953291535377502,
|
| 1589 |
+
"learning_rate": 1.604742930767298e-05,
|
| 1590 |
+
"loss": 0.0219,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 4.149908592321755,
|
| 1595 |
+
"grad_norm": 0.18694134056568146,
|
| 1596 |
+
"learning_rate": 1.5396581677124124e-05,
|
| 1597 |
+
"loss": 0.0169,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 4.16819012797075,
|
| 1602 |
+
"grad_norm": 0.17348328232765198,
|
| 1603 |
+
"learning_rate": 1.4758105025599068e-05,
|
| 1604 |
+
"loss": 0.0159,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 4.186471663619744,
|
| 1609 |
+
"grad_norm": 0.16517849266529083,
|
| 1610 |
+
"learning_rate": 1.4132092713170242e-05,
|
| 1611 |
+
"loss": 0.0137,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 4.204753199268739,
|
| 1616 |
+
"grad_norm": 0.13645470142364502,
|
| 1617 |
+
"learning_rate": 1.3518636277335084e-05,
|
| 1618 |
+
"loss": 0.0149,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 4.223034734917733,
|
| 1623 |
+
"grad_norm": 0.14027458429336548,
|
| 1624 |
+
"learning_rate": 1.291782541963107e-05,
|
| 1625 |
+
"loss": 0.0147,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 4.2413162705667276,
|
| 1630 |
+
"grad_norm": 0.11632464081048965,
|
| 1631 |
+
"learning_rate": 1.2329747992519269e-05,
|
| 1632 |
+
"loss": 0.0137,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 4.259597806215722,
|
| 1637 |
+
"grad_norm": 0.2426212579011917,
|
| 1638 |
+
"learning_rate": 1.1754489986538419e-05,
|
| 1639 |
+
"loss": 0.0117,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 4.2778793418647165,
|
| 1644 |
+
"grad_norm": 0.20155277848243713,
|
| 1645 |
+
"learning_rate": 1.1192135517730884e-05,
|
| 1646 |
+
"loss": 0.0147,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 4.296160877513711,
|
| 1651 |
+
"grad_norm": 0.14590322971343994,
|
| 1652 |
+
"learning_rate": 1.0642766815343196e-05,
|
| 1653 |
+
"loss": 0.0119,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 4.3144424131627055,
|
| 1658 |
+
"grad_norm": 0.17194287478923798,
|
| 1659 |
+
"learning_rate": 1.0106464209802013e-05,
|
| 1660 |
+
"loss": 0.0115,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 4.3327239488117,
|
| 1665 |
+
"grad_norm": 0.243038609623909,
|
| 1666 |
+
"learning_rate": 9.583306120968072e-06,
|
| 1667 |
+
"loss": 0.0153,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 4.3510054844606945,
|
| 1672 |
+
"grad_norm": 0.29729005694389343,
|
| 1673 |
+
"learning_rate": 9.0733690466694e-06,
|
| 1674 |
+
"loss": 0.0136,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 4.369287020109689,
|
| 1679 |
+
"grad_norm": 0.1595577597618103,
|
| 1680 |
+
"learning_rate": 8.576727551515474e-06,
|
| 1681 |
+
"loss": 0.0156,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 4.387568555758683,
|
| 1686 |
+
"grad_norm": 0.12783007323741913,
|
| 1687 |
+
"learning_rate": 8.093454255994248e-06,
|
| 1688 |
+
"loss": 0.0122,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 4.405850091407678,
|
| 1693 |
+
"grad_norm": 0.26608356833457947,
|
| 1694 |
+
"learning_rate": 7.6236198258532675e-06,
|
| 1695 |
+
"loss": 0.0136,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 4.424131627056672,
|
| 1700 |
+
"grad_norm": 0.1889527142047882,
|
| 1701 |
+
"learning_rate": 7.167292961766725e-06,
|
| 1702 |
+
"loss": 0.015,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 4.442413162705667,
|
| 1707 |
+
"grad_norm": 0.2580418884754181,
|
| 1708 |
+
"learning_rate": 6.724540389289913e-06,
|
| 1709 |
+
"loss": 0.0132,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 4.460694698354661,
|
| 1714 |
+
"grad_norm": 0.22082190215587616,
|
| 1715 |
+
"learning_rate": 6.295426849102271e-06,
|
| 1716 |
+
"loss": 0.0113,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 4.478976234003657,
|
| 1721 |
+
"grad_norm": 0.11176195740699768,
|
| 1722 |
+
"learning_rate": 5.8800150875408574e-06,
|
| 1723 |
+
"loss": 0.0141,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 4.497257769652651,
|
| 1728 |
+
"grad_norm": 0.1779015064239502,
|
| 1729 |
+
"learning_rate": 5.478365847425449e-06,
|
| 1730 |
+
"loss": 0.0113,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 4.515539305301646,
|
| 1735 |
+
"grad_norm": 0.15661382675170898,
|
| 1736 |
+
"learning_rate": 5.090537859176425e-06,
|
| 1737 |
+
"loss": 0.0102,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 4.53382084095064,
|
| 1742 |
+
"grad_norm": 0.21932142972946167,
|
| 1743 |
+
"learning_rate": 4.716587832227071e-06,
|
| 1744 |
+
"loss": 0.0147,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 4.552102376599635,
|
| 1749 |
+
"grad_norm": 0.30200353264808655,
|
| 1750 |
+
"learning_rate": 4.356570446731356e-06,
|
| 1751 |
+
"loss": 0.0152,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 4.570383912248629,
|
| 1756 |
+
"grad_norm": 0.11431296914815903,
|
| 1757 |
+
"learning_rate": 4.010538345568371e-06,
|
| 1758 |
+
"loss": 0.017,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 4.588665447897624,
|
| 1763 |
+
"grad_norm": 0.2187824845314026,
|
| 1764 |
+
"learning_rate": 3.678542126644813e-06,
|
| 1765 |
+
"loss": 0.0168,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 4.606946983546618,
|
| 1770 |
+
"grad_norm": 0.12425347417593002,
|
| 1771 |
+
"learning_rate": 3.360630335496362e-06,
|
| 1772 |
+
"loss": 0.0113,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 4.625228519195613,
|
| 1777 |
+
"grad_norm": 0.17450736463069916,
|
| 1778 |
+
"learning_rate": 3.056849458189115e-06,
|
| 1779 |
+
"loss": 0.015,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 4.643510054844607,
|
| 1784 |
+
"grad_norm": 0.2220509946346283,
|
| 1785 |
+
"learning_rate": 2.7672439145223773e-06,
|
| 1786 |
+
"loss": 0.0196,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 4.661791590493602,
|
| 1791 |
+
"grad_norm": 0.2917903959751129,
|
| 1792 |
+
"learning_rate": 2.491856051533392e-06,
|
| 1793 |
+
"loss": 0.0165,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 4.680073126142596,
|
| 1798 |
+
"grad_norm": 0.22880949079990387,
|
| 1799 |
+
"learning_rate": 2.230726137305206e-06,
|
| 1800 |
+
"loss": 0.0165,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 4.698354661791591,
|
| 1805 |
+
"grad_norm": 0.2307160645723343,
|
| 1806 |
+
"learning_rate": 1.983892355078587e-06,
|
| 1807 |
+
"loss": 0.0129,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 4.716636197440585,
|
| 1812 |
+
"grad_norm": 0.1975175142288208,
|
| 1813 |
+
"learning_rate": 1.7513907976687283e-06,
|
| 1814 |
+
"loss": 0.016,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 4.7349177330895795,
|
| 1819 |
+
"grad_norm": 0.23436793684959412,
|
| 1820 |
+
"learning_rate": 1.533255462187666e-06,
|
| 1821 |
+
"loss": 0.0108,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 4.753199268738574,
|
| 1826 |
+
"grad_norm": 0.14805355668067932,
|
| 1827 |
+
"learning_rate": 1.329518245073047e-06,
|
| 1828 |
+
"loss": 0.0182,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 4.7714808043875685,
|
| 1833 |
+
"grad_norm": 0.1988326609134674,
|
| 1834 |
+
"learning_rate": 1.1402089374242365e-06,
|
| 1835 |
+
"loss": 0.0119,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 4.789762340036563,
|
| 1840 |
+
"grad_norm": 0.12207505851984024,
|
| 1841 |
+
"learning_rate": 9.65355220646036e-07,
|
| 1842 |
+
"loss": 0.0128,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 4.8080438756855575,
|
| 1847 |
+
"grad_norm": 0.1775001883506775,
|
| 1848 |
+
"learning_rate": 8.049826624011881e-07,
|
| 1849 |
+
"loss": 0.0166,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 4.826325411334552,
|
| 1854 |
+
"grad_norm": 0.2577812075614929,
|
| 1855 |
+
"learning_rate": 6.591147128716224e-07,
|
| 1856 |
+
"loss": 0.0191,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 4.844606946983546,
|
| 1861 |
+
"grad_norm": 0.1870380938053131,
|
| 1862 |
+
"learning_rate": 5.277727013296097e-07,
|
| 1863 |
+
"loss": 0.0125,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 4.862888482632541,
|
| 1868 |
+
"grad_norm": 0.22090613842010498,
|
| 1869 |
+
"learning_rate": 4.1097583301888954e-07,
|
| 1870 |
+
"loss": 0.009,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 4.881170018281535,
|
| 1875 |
+
"grad_norm": 0.25381821393966675,
|
| 1876 |
+
"learning_rate": 3.0874118634640626e-07,
|
| 1877 |
+
"loss": 0.0158,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 4.89945155393053,
|
| 1882 |
+
"grad_norm": 0.25577688217163086,
|
| 1883 |
+
"learning_rate": 2.210837103850949e-07,
|
| 1884 |
+
"loss": 0.0074,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 4.917733089579524,
|
| 1889 |
+
"grad_norm": 0.1378592997789383,
|
| 1890 |
+
"learning_rate": 1.4801622268791892e-07,
|
| 1891 |
+
"loss": 0.0104,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 4.936014625228519,
|
| 1896 |
+
"grad_norm": 0.1672651469707489,
|
| 1897 |
+
"learning_rate": 8.954940741369155e-08,
|
| 1898 |
+
"loss": 0.0126,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 4.954296160877513,
|
| 1903 |
+
"grad_norm": 0.10131768137216568,
|
| 1904 |
+
"learning_rate": 4.5691813764803247e-08,
|
| 1905 |
+
"loss": 0.0093,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 4.972577696526509,
|
| 1910 |
+
"grad_norm": 0.19686748087406158,
|
| 1911 |
+
"learning_rate": 1.644985473709948e-08,
|
| 1912 |
+
"loss": 0.0132,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 4.990859232175502,
|
| 1917 |
+
"grad_norm": 0.16079658269882202,
|
| 1918 |
+
"learning_rate": 1.8278061821863646e-09,
|
| 1919 |
+
"loss": 0.0096,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 5.0,
|
| 1924 |
+
"step": 2735,
|
| 1925 |
+
"total_flos": 9.752547304210464e+16,
|
| 1926 |
+
"train_loss": 0.045771664261164136,
|
| 1927 |
+
"train_runtime": 1237.486,
|
| 1928 |
+
"train_samples_per_second": 35.362,
|
| 1929 |
+
"train_steps_per_second": 2.21
|
| 1930 |
+
}
|
| 1931 |
+
],
|
| 1932 |
+
"logging_steps": 10,
|
| 1933 |
+
"max_steps": 2735,
|
| 1934 |
+
"num_input_tokens_seen": 0,
|
| 1935 |
+
"num_train_epochs": 5,
|
| 1936 |
+
"save_steps": 500,
|
| 1937 |
+
"stateful_callbacks": {
|
| 1938 |
+
"TrainerControl": {
|
| 1939 |
+
"args": {
|
| 1940 |
+
"should_epoch_stop": false,
|
| 1941 |
+
"should_evaluate": false,
|
| 1942 |
+
"should_log": false,
|
| 1943 |
+
"should_save": true,
|
| 1944 |
+
"should_training_stop": true
|
| 1945 |
+
},
|
| 1946 |
+
"attributes": {}
|
| 1947 |
+
}
|
| 1948 |
+
},
|
| 1949 |
+
"total_flos": 9.752547304210464e+16,
|
| 1950 |
+
"train_batch_size": 16,
|
| 1951 |
+
"trial_name": null,
|
| 1952 |
+
"trial_params": null
|
| 1953 |
+
}
|