Upload multi_student_2/checkpoint-3600/trainer_state.json with huggingface_hub
Browse files
multi_student_2/checkpoint-3600/trainer_state.json
ADDED
|
@@ -0,0 +1,2554 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.8552161840539467,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3600,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.007933359777865925,
|
| 14 |
+
"grad_norm": 19.14600299430099,
|
| 15 |
+
"learning_rate": 8.9197224975223e-07,
|
| 16 |
+
"loss": 8.2489,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.01586671955573185,
|
| 21 |
+
"grad_norm": 18.01915458135243,
|
| 22 |
+
"learning_rate": 1.8830525272547078e-06,
|
| 23 |
+
"loss": 8.1177,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.02380007933359778,
|
| 28 |
+
"grad_norm": 9.33822311099433,
|
| 29 |
+
"learning_rate": 2.8741328047571853e-06,
|
| 30 |
+
"loss": 7.4735,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.0317334391114637,
|
| 35 |
+
"grad_norm": 4.906820841029322,
|
| 36 |
+
"learning_rate": 3.8652130822596635e-06,
|
| 37 |
+
"loss": 6.7427,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.039666798889329634,
|
| 42 |
+
"grad_norm": 3.0516622673060785,
|
| 43 |
+
"learning_rate": 4.8562933597621405e-06,
|
| 44 |
+
"loss": 6.2707,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.04760015866719556,
|
| 49 |
+
"grad_norm": 3.0497788345357493,
|
| 50 |
+
"learning_rate": 5.847373637264618e-06,
|
| 51 |
+
"loss": 5.9319,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.055533518445061485,
|
| 56 |
+
"grad_norm": 3.678614138586657,
|
| 57 |
+
"learning_rate": 6.838453914767097e-06,
|
| 58 |
+
"loss": 5.5292,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.0634668782229274,
|
| 63 |
+
"grad_norm": 4.261516409919679,
|
| 64 |
+
"learning_rate": 7.829534192269576e-06,
|
| 65 |
+
"loss": 4.9099,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.07140023800079334,
|
| 70 |
+
"grad_norm": 4.6638020698067875,
|
| 71 |
+
"learning_rate": 8.820614469772052e-06,
|
| 72 |
+
"loss": 4.2313,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.07933359777865927,
|
| 77 |
+
"grad_norm": 4.5951745742169,
|
| 78 |
+
"learning_rate": 9.81169474727453e-06,
|
| 79 |
+
"loss": 3.663,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.0872669575565252,
|
| 84 |
+
"grad_norm": 3.9132985800446947,
|
| 85 |
+
"learning_rate": 1.0802775024777007e-05,
|
| 86 |
+
"loss": 3.2806,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.09520031733439112,
|
| 91 |
+
"grad_norm": 3.864223439251188,
|
| 92 |
+
"learning_rate": 1.1793855302279485e-05,
|
| 93 |
+
"loss": 3.0464,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.10313367711225704,
|
| 98 |
+
"grad_norm": 3.5022825994548232,
|
| 99 |
+
"learning_rate": 1.2784935579781963e-05,
|
| 100 |
+
"loss": 2.8886,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.11106703689012297,
|
| 105 |
+
"grad_norm": 5.1510905388003,
|
| 106 |
+
"learning_rate": 1.377601585728444e-05,
|
| 107 |
+
"loss": 2.7668,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.1190003966679889,
|
| 112 |
+
"grad_norm": 7.21771904799698,
|
| 113 |
+
"learning_rate": 1.4767096134786917e-05,
|
| 114 |
+
"loss": 2.6692,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.1269337564458548,
|
| 119 |
+
"grad_norm": 9.497867401163038,
|
| 120 |
+
"learning_rate": 1.5758176412289396e-05,
|
| 121 |
+
"loss": 2.5934,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.13486711622372075,
|
| 126 |
+
"grad_norm": 8.479404348386907,
|
| 127 |
+
"learning_rate": 1.6749256689791874e-05,
|
| 128 |
+
"loss": 2.5471,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.14280047600158668,
|
| 133 |
+
"grad_norm": 13.33424723235721,
|
| 134 |
+
"learning_rate": 1.7740336967294352e-05,
|
| 135 |
+
"loss": 2.4827,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.1507338357794526,
|
| 140 |
+
"grad_norm": 9.736118445504315,
|
| 141 |
+
"learning_rate": 1.873141724479683e-05,
|
| 142 |
+
"loss": 2.4223,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.15866719555731854,
|
| 147 |
+
"grad_norm": 12.548385792238333,
|
| 148 |
+
"learning_rate": 1.9722497522299307e-05,
|
| 149 |
+
"loss": 2.3752,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.16660055533518445,
|
| 154 |
+
"grad_norm": 15.910150113440148,
|
| 155 |
+
"learning_rate": 2.0713577799801785e-05,
|
| 156 |
+
"loss": 2.3339,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.1745339151130504,
|
| 161 |
+
"grad_norm": 12.024121798784853,
|
| 162 |
+
"learning_rate": 2.1704658077304263e-05,
|
| 163 |
+
"loss": 2.2732,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.1824672748909163,
|
| 168 |
+
"grad_norm": 18.143793557621525,
|
| 169 |
+
"learning_rate": 2.269573835480674e-05,
|
| 170 |
+
"loss": 2.2154,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.19040063466878224,
|
| 175 |
+
"grad_norm": 11.930035885996817,
|
| 176 |
+
"learning_rate": 2.368681863230922e-05,
|
| 177 |
+
"loss": 2.1602,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.19833399444664815,
|
| 182 |
+
"grad_norm": 12.424586861990159,
|
| 183 |
+
"learning_rate": 2.4677898909811696e-05,
|
| 184 |
+
"loss": 2.0887,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.2062673542245141,
|
| 189 |
+
"grad_norm": 11.063682700121975,
|
| 190 |
+
"learning_rate": 2.5668979187314174e-05,
|
| 191 |
+
"loss": 1.9843,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.21420071400238,
|
| 196 |
+
"grad_norm": 9.80005982816964,
|
| 197 |
+
"learning_rate": 2.6660059464816652e-05,
|
| 198 |
+
"loss": 1.871,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.22213407378024594,
|
| 203 |
+
"grad_norm": 6.8486046511969985,
|
| 204 |
+
"learning_rate": 2.7651139742319126e-05,
|
| 205 |
+
"loss": 1.7654,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.23006743355811185,
|
| 210 |
+
"grad_norm": 4.650133260417432,
|
| 211 |
+
"learning_rate": 2.8642220019821607e-05,
|
| 212 |
+
"loss": 1.6631,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.2380007933359778,
|
| 217 |
+
"grad_norm": 8.484902376448611,
|
| 218 |
+
"learning_rate": 2.9633300297324085e-05,
|
| 219 |
+
"loss": 1.587,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.2459341531138437,
|
| 224 |
+
"grad_norm": 8.266296207037957,
|
| 225 |
+
"learning_rate": 3.062438057482656e-05,
|
| 226 |
+
"loss": 1.5209,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.2538675128917096,
|
| 231 |
+
"grad_norm": 6.035258103391185,
|
| 232 |
+
"learning_rate": 3.161546085232904e-05,
|
| 233 |
+
"loss": 1.4691,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.2618008726695756,
|
| 238 |
+
"grad_norm": 6.090335574638085,
|
| 239 |
+
"learning_rate": 3.260654112983152e-05,
|
| 240 |
+
"loss": 1.4213,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.2697342324474415,
|
| 245 |
+
"grad_norm": 7.504625709152101,
|
| 246 |
+
"learning_rate": 3.359762140733399e-05,
|
| 247 |
+
"loss": 1.3809,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.2776675922253074,
|
| 252 |
+
"grad_norm": 5.216278847793629,
|
| 253 |
+
"learning_rate": 3.458870168483647e-05,
|
| 254 |
+
"loss": 1.3304,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.28560095200317337,
|
| 259 |
+
"grad_norm": 5.001664767324889,
|
| 260 |
+
"learning_rate": 3.5579781962338955e-05,
|
| 261 |
+
"loss": 1.2886,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.2935343117810393,
|
| 266 |
+
"grad_norm": 4.035122456050597,
|
| 267 |
+
"learning_rate": 3.657086223984143e-05,
|
| 268 |
+
"loss": 1.2441,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.3014676715589052,
|
| 273 |
+
"grad_norm": 4.3131477681191015,
|
| 274 |
+
"learning_rate": 3.7561942517343904e-05,
|
| 275 |
+
"loss": 1.1973,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.3094010313367711,
|
| 280 |
+
"grad_norm": 8.188203831814278,
|
| 281 |
+
"learning_rate": 3.855302279484638e-05,
|
| 282 |
+
"loss": 1.175,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.31733439111463707,
|
| 287 |
+
"grad_norm": 4.755933403654607,
|
| 288 |
+
"learning_rate": 3.9544103072348866e-05,
|
| 289 |
+
"loss": 1.1342,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.325267750892503,
|
| 294 |
+
"grad_norm": 6.57796715979494,
|
| 295 |
+
"learning_rate": 4.0535183349851344e-05,
|
| 296 |
+
"loss": 1.0971,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.3332011106703689,
|
| 301 |
+
"grad_norm": 4.216167153902993,
|
| 302 |
+
"learning_rate": 4.1526263627353815e-05,
|
| 303 |
+
"loss": 1.0513,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.3411344704482348,
|
| 308 |
+
"grad_norm": 8.418073662598946,
|
| 309 |
+
"learning_rate": 4.251734390485629e-05,
|
| 310 |
+
"loss": 1.0278,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.3490678302261008,
|
| 315 |
+
"grad_norm": 4.293832971560352,
|
| 316 |
+
"learning_rate": 4.350842418235878e-05,
|
| 317 |
+
"loss": 0.9989,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.3570011900039667,
|
| 322 |
+
"grad_norm": 4.738542701520305,
|
| 323 |
+
"learning_rate": 4.4499504459861255e-05,
|
| 324 |
+
"loss": 0.9646,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.3649345497818326,
|
| 329 |
+
"grad_norm": 3.836235781620353,
|
| 330 |
+
"learning_rate": 4.5490584737363726e-05,
|
| 331 |
+
"loss": 0.9282,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.3728679095596985,
|
| 336 |
+
"grad_norm": 4.385628595929679,
|
| 337 |
+
"learning_rate": 4.6481665014866204e-05,
|
| 338 |
+
"loss": 0.9069,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.3808012693375645,
|
| 343 |
+
"grad_norm": 4.795190260431852,
|
| 344 |
+
"learning_rate": 4.747274529236869e-05,
|
| 345 |
+
"loss": 0.8767,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.3887346291154304,
|
| 350 |
+
"grad_norm": 5.15771361003216,
|
| 351 |
+
"learning_rate": 4.846382556987116e-05,
|
| 352 |
+
"loss": 0.848,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.3966679888932963,
|
| 357 |
+
"grad_norm": 5.00866147389186,
|
| 358 |
+
"learning_rate": 4.945490584737364e-05,
|
| 359 |
+
"loss": 0.832,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.40460134867116226,
|
| 364 |
+
"grad_norm": 3.3749478011742617,
|
| 365 |
+
"learning_rate": 5.044598612487612e-05,
|
| 366 |
+
"loss": 0.801,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.4125347084490282,
|
| 371 |
+
"grad_norm": 5.648100107599949,
|
| 372 |
+
"learning_rate": 5.14370664023786e-05,
|
| 373 |
+
"loss": 0.7907,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.4204680682268941,
|
| 378 |
+
"grad_norm": 5.180183508710987,
|
| 379 |
+
"learning_rate": 5.242814667988107e-05,
|
| 380 |
+
"loss": 0.7641,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.42840142800476,
|
| 385 |
+
"grad_norm": 4.795683231567134,
|
| 386 |
+
"learning_rate": 5.341922695738355e-05,
|
| 387 |
+
"loss": 0.7428,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.43633478778262597,
|
| 392 |
+
"grad_norm": 4.474423469358904,
|
| 393 |
+
"learning_rate": 5.4410307234886027e-05,
|
| 394 |
+
"loss": 0.7178,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.4442681475604919,
|
| 399 |
+
"grad_norm": 5.965818045702901,
|
| 400 |
+
"learning_rate": 5.5401387512388504e-05,
|
| 401 |
+
"loss": 0.7008,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.4522015073383578,
|
| 406 |
+
"grad_norm": 4.4181008640757256,
|
| 407 |
+
"learning_rate": 5.639246778989098e-05,
|
| 408 |
+
"loss": 0.678,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.4601348671162237,
|
| 413 |
+
"grad_norm": 4.925348313364693,
|
| 414 |
+
"learning_rate": 5.738354806739347e-05,
|
| 415 |
+
"loss": 0.6516,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.46806822689408967,
|
| 420 |
+
"grad_norm": 4.995771437156914,
|
| 421 |
+
"learning_rate": 5.8374628344895944e-05,
|
| 422 |
+
"loss": 0.6261,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.4760015866719556,
|
| 427 |
+
"grad_norm": 4.3400967454545185,
|
| 428 |
+
"learning_rate": 5.936570862239842e-05,
|
| 429 |
+
"loss": 0.6099,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.4839349464498215,
|
| 434 |
+
"grad_norm": 3.8437851403940013,
|
| 435 |
+
"learning_rate": 6.035678889990089e-05,
|
| 436 |
+
"loss": 0.5905,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.4918683062276874,
|
| 441 |
+
"grad_norm": 4.673291642824443,
|
| 442 |
+
"learning_rate": 6.134786917740337e-05,
|
| 443 |
+
"loss": 0.5739,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.49980166600555337,
|
| 448 |
+
"grad_norm": 4.691220121981611,
|
| 449 |
+
"learning_rate": 6.233894945490585e-05,
|
| 450 |
+
"loss": 0.5543,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.5077350257834192,
|
| 455 |
+
"grad_norm": 4.609768045302665,
|
| 456 |
+
"learning_rate": 6.333002973240833e-05,
|
| 457 |
+
"loss": 0.5265,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.5156683855612852,
|
| 462 |
+
"grad_norm": 3.2621270260578967,
|
| 463 |
+
"learning_rate": 6.43211100099108e-05,
|
| 464 |
+
"loss": 0.5175,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.5236017453391512,
|
| 469 |
+
"grad_norm": 3.5703471859456806,
|
| 470 |
+
"learning_rate": 6.531219028741328e-05,
|
| 471 |
+
"loss": 0.4972,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.5315351051170171,
|
| 476 |
+
"grad_norm": 3.530519835264167,
|
| 477 |
+
"learning_rate": 6.630327056491576e-05,
|
| 478 |
+
"loss": 0.4717,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.539468464894883,
|
| 483 |
+
"grad_norm": 2.392046097161058,
|
| 484 |
+
"learning_rate": 6.729435084241824e-05,
|
| 485 |
+
"loss": 0.4583,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.5474018246727489,
|
| 490 |
+
"grad_norm": 3.149236166223247,
|
| 491 |
+
"learning_rate": 6.828543111992072e-05,
|
| 492 |
+
"loss": 0.4412,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.5553351844506148,
|
| 497 |
+
"grad_norm": 2.9354688050931586,
|
| 498 |
+
"learning_rate": 6.92765113974232e-05,
|
| 499 |
+
"loss": 0.423,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.5632685442284807,
|
| 504 |
+
"grad_norm": 2.7127436235307125,
|
| 505 |
+
"learning_rate": 7.026759167492567e-05,
|
| 506 |
+
"loss": 0.4062,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.5712019040063467,
|
| 511 |
+
"grad_norm": 2.6410601628390777,
|
| 512 |
+
"learning_rate": 7.125867195242815e-05,
|
| 513 |
+
"loss": 0.3923,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.5791352637842127,
|
| 518 |
+
"grad_norm": 2.7996527641631386,
|
| 519 |
+
"learning_rate": 7.224975222993063e-05,
|
| 520 |
+
"loss": 0.3739,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.5870686235620786,
|
| 525 |
+
"grad_norm": 3.53218129873484,
|
| 526 |
+
"learning_rate": 7.32408325074331e-05,
|
| 527 |
+
"loss": 0.3558,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.5950019833399445,
|
| 532 |
+
"grad_norm": 3.256471177277615,
|
| 533 |
+
"learning_rate": 7.423191278493558e-05,
|
| 534 |
+
"loss": 0.3378,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.6029353431178104,
|
| 539 |
+
"grad_norm": 2.448855772936633,
|
| 540 |
+
"learning_rate": 7.522299306243806e-05,
|
| 541 |
+
"loss": 0.3318,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.6108687028956763,
|
| 546 |
+
"grad_norm": 2.8206966008765355,
|
| 547 |
+
"learning_rate": 7.621407333994054e-05,
|
| 548 |
+
"loss": 0.3145,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.6188020626735422,
|
| 553 |
+
"grad_norm": 2.262400615934418,
|
| 554 |
+
"learning_rate": 7.720515361744302e-05,
|
| 555 |
+
"loss": 0.2981,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.6267354224514081,
|
| 560 |
+
"grad_norm": 2.6691511066584894,
|
| 561 |
+
"learning_rate": 7.81962338949455e-05,
|
| 562 |
+
"loss": 0.2878,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.6346687822292741,
|
| 567 |
+
"grad_norm": 2.1502320771866494,
|
| 568 |
+
"learning_rate": 7.918731417244797e-05,
|
| 569 |
+
"loss": 0.2724,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.64260214200714,
|
| 574 |
+
"grad_norm": 2.1869290626436806,
|
| 575 |
+
"learning_rate": 8.017839444995045e-05,
|
| 576 |
+
"loss": 0.2589,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.650535501785006,
|
| 581 |
+
"grad_norm": 2.6277829837493476,
|
| 582 |
+
"learning_rate": 8.116947472745293e-05,
|
| 583 |
+
"loss": 0.2481,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.6584688615628719,
|
| 588 |
+
"grad_norm": 1.8401496589236042,
|
| 589 |
+
"learning_rate": 8.21605550049554e-05,
|
| 590 |
+
"loss": 0.2361,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.6664022213407378,
|
| 595 |
+
"grad_norm": 1.7862155383026863,
|
| 596 |
+
"learning_rate": 8.315163528245788e-05,
|
| 597 |
+
"loss": 0.222,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.6743355811186037,
|
| 602 |
+
"grad_norm": 1.9765212410305726,
|
| 603 |
+
"learning_rate": 8.414271555996036e-05,
|
| 604 |
+
"loss": 0.2103,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.6822689408964696,
|
| 609 |
+
"grad_norm": 2.4397348134664556,
|
| 610 |
+
"learning_rate": 8.513379583746284e-05,
|
| 611 |
+
"loss": 0.2021,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.6902023006743356,
|
| 616 |
+
"grad_norm": 2.296758487647809,
|
| 617 |
+
"learning_rate": 8.612487611496532e-05,
|
| 618 |
+
"loss": 0.1924,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.6981356604522015,
|
| 623 |
+
"grad_norm": 2.104859870669743,
|
| 624 |
+
"learning_rate": 8.71159563924678e-05,
|
| 625 |
+
"loss": 0.1817,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.7060690202300675,
|
| 630 |
+
"grad_norm": 1.3003724019670297,
|
| 631 |
+
"learning_rate": 8.810703666997027e-05,
|
| 632 |
+
"loss": 0.1732,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.7140023800079334,
|
| 637 |
+
"grad_norm": 1.3709932626535646,
|
| 638 |
+
"learning_rate": 8.909811694747275e-05,
|
| 639 |
+
"loss": 0.1593,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.7219357397857993,
|
| 644 |
+
"grad_norm": 1.8342489151171628,
|
| 645 |
+
"learning_rate": 9.008919722497523e-05,
|
| 646 |
+
"loss": 0.1551,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.7298690995636652,
|
| 651 |
+
"grad_norm": 1.762934595290678,
|
| 652 |
+
"learning_rate": 9.10802775024777e-05,
|
| 653 |
+
"loss": 0.1489,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.7378024593415311,
|
| 658 |
+
"grad_norm": 1.4075349928757257,
|
| 659 |
+
"learning_rate": 9.207135777998018e-05,
|
| 660 |
+
"loss": 0.1391,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.745735819119397,
|
| 665 |
+
"grad_norm": 1.3962752805987728,
|
| 666 |
+
"learning_rate": 9.306243805748266e-05,
|
| 667 |
+
"loss": 0.1332,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.753669178897263,
|
| 672 |
+
"grad_norm": 3.5595641910775173,
|
| 673 |
+
"learning_rate": 9.405351833498514e-05,
|
| 674 |
+
"loss": 0.1507,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.761602538675129,
|
| 679 |
+
"grad_norm": 1.2799346026465042,
|
| 680 |
+
"learning_rate": 9.504459861248762e-05,
|
| 681 |
+
"loss": 0.1706,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.7695358984529949,
|
| 686 |
+
"grad_norm": 0.9866434355012127,
|
| 687 |
+
"learning_rate": 9.60356788899901e-05,
|
| 688 |
+
"loss": 0.1241,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.7774692582308608,
|
| 693 |
+
"grad_norm": 1.2605525685265588,
|
| 694 |
+
"learning_rate": 9.702675916749257e-05,
|
| 695 |
+
"loss": 0.1133,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.7854026180087267,
|
| 700 |
+
"grad_norm": 1.0654588120046724,
|
| 701 |
+
"learning_rate": 9.801783944499505e-05,
|
| 702 |
+
"loss": 0.1066,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.7933359777865926,
|
| 707 |
+
"grad_norm": 0.761881486606698,
|
| 708 |
+
"learning_rate": 9.900891972249753e-05,
|
| 709 |
+
"loss": 0.1015,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.8012693375644585,
|
| 714 |
+
"grad_norm": 1.0601392633604383,
|
| 715 |
+
"learning_rate": 0.0001,
|
| 716 |
+
"loss": 0.0953,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.8092026973423245,
|
| 721 |
+
"grad_norm": 0.8224204361157935,
|
| 722 |
+
"learning_rate": 9.999970066101032e-05,
|
| 723 |
+
"loss": 0.0906,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.8171360571201904,
|
| 728 |
+
"grad_norm": 0.7080480512831684,
|
| 729 |
+
"learning_rate": 9.99988026476254e-05,
|
| 730 |
+
"loss": 0.0866,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.8250694168980564,
|
| 735 |
+
"grad_norm": 0.7547638843757503,
|
| 736 |
+
"learning_rate": 9.999730597059764e-05,
|
| 737 |
+
"loss": 0.0809,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.8330027766759223,
|
| 742 |
+
"grad_norm": 1.1588422105514353,
|
| 743 |
+
"learning_rate": 9.999521064784764e-05,
|
| 744 |
+
"loss": 0.0844,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.8409361364537882,
|
| 749 |
+
"grad_norm": 0.6385523334282596,
|
| 750 |
+
"learning_rate": 9.999251670446386e-05,
|
| 751 |
+
"loss": 0.0804,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.8488694962316541,
|
| 756 |
+
"grad_norm": 0.5623095327146973,
|
| 757 |
+
"learning_rate": 9.998922417270235e-05,
|
| 758 |
+
"loss": 0.072,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.85680285600952,
|
| 763 |
+
"grad_norm": 0.43678572122080983,
|
| 764 |
+
"learning_rate": 9.99853330919865e-05,
|
| 765 |
+
"loss": 0.0679,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.8647362157873859,
|
| 770 |
+
"grad_norm": 0.4972727247312438,
|
| 771 |
+
"learning_rate": 9.998084350890633e-05,
|
| 772 |
+
"loss": 0.065,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.8726695755652519,
|
| 777 |
+
"grad_norm": 0.5035917631799433,
|
| 778 |
+
"learning_rate": 9.997575547721818e-05,
|
| 779 |
+
"loss": 0.0627,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.8806029353431178,
|
| 784 |
+
"grad_norm": 0.5182605794595788,
|
| 785 |
+
"learning_rate": 9.997006905784388e-05,
|
| 786 |
+
"loss": 0.0602,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.8885362951209838,
|
| 791 |
+
"grad_norm": 0.41920699625736624,
|
| 792 |
+
"learning_rate": 9.996378431887012e-05,
|
| 793 |
+
"loss": 0.0579,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.8964696548988497,
|
| 798 |
+
"grad_norm": 0.414907297365465,
|
| 799 |
+
"learning_rate": 9.995690133554758e-05,
|
| 800 |
+
"loss": 0.0546,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.9044030146767156,
|
| 805 |
+
"grad_norm": 0.3702005062489374,
|
| 806 |
+
"learning_rate": 9.99494201902901e-05,
|
| 807 |
+
"loss": 0.0523,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.9123363744545815,
|
| 812 |
+
"grad_norm": 0.42037834689750175,
|
| 813 |
+
"learning_rate": 9.994134097267358e-05,
|
| 814 |
+
"loss": 0.0516,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.9202697342324474,
|
| 819 |
+
"grad_norm": 0.38928699328314564,
|
| 820 |
+
"learning_rate": 9.993266377943504e-05,
|
| 821 |
+
"loss": 0.0505,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.9282030940103134,
|
| 826 |
+
"grad_norm": 0.3941895011795836,
|
| 827 |
+
"learning_rate": 9.992338871447138e-05,
|
| 828 |
+
"loss": 0.0495,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.9361364537881793,
|
| 833 |
+
"grad_norm": 0.37452365492894446,
|
| 834 |
+
"learning_rate": 9.991351588883811e-05,
|
| 835 |
+
"loss": 0.0474,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.9440698135660452,
|
| 840 |
+
"grad_norm": 0.4395025405712049,
|
| 841 |
+
"learning_rate": 9.99030454207481e-05,
|
| 842 |
+
"loss": 0.0467,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.9520031733439112,
|
| 847 |
+
"grad_norm": 0.4639716702934286,
|
| 848 |
+
"learning_rate": 9.989197743557017e-05,
|
| 849 |
+
"loss": 0.045,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.9599365331217771,
|
| 854 |
+
"grad_norm": 0.34908102760028864,
|
| 855 |
+
"learning_rate": 9.988031206582746e-05,
|
| 856 |
+
"loss": 0.0439,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.967869892899643,
|
| 861 |
+
"grad_norm": 0.29211915294831964,
|
| 862 |
+
"learning_rate": 9.986804945119597e-05,
|
| 863 |
+
"loss": 0.0429,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.9758032526775089,
|
| 868 |
+
"grad_norm": 0.2801643208729514,
|
| 869 |
+
"learning_rate": 9.985518973850285e-05,
|
| 870 |
+
"loss": 0.041,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.9837366124553748,
|
| 875 |
+
"grad_norm": 0.3013251709022931,
|
| 876 |
+
"learning_rate": 9.984173308172465e-05,
|
| 877 |
+
"loss": 0.0403,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.9916699722332408,
|
| 882 |
+
"grad_norm": 0.28830525385588507,
|
| 883 |
+
"learning_rate": 9.982767964198542e-05,
|
| 884 |
+
"loss": 0.04,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.9996033320111067,
|
| 889 |
+
"grad_norm": 0.2756668810883745,
|
| 890 |
+
"learning_rate": 9.98130295875549e-05,
|
| 891 |
+
"loss": 0.0386,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.0071400238000794,
|
| 896 |
+
"grad_norm": 0.2673295493921112,
|
| 897 |
+
"learning_rate": 9.979778309384638e-05,
|
| 898 |
+
"loss": 0.0369,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.0150733835779453,
|
| 903 |
+
"grad_norm": 0.28881709817045276,
|
| 904 |
+
"learning_rate": 9.978194034341462e-05,
|
| 905 |
+
"loss": 0.0366,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.0230067433558112,
|
| 910 |
+
"grad_norm": 0.2443064331996704,
|
| 911 |
+
"learning_rate": 9.976550152595379e-05,
|
| 912 |
+
"loss": 0.035,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.0309401031336771,
|
| 917 |
+
"grad_norm": 0.2586793532004466,
|
| 918 |
+
"learning_rate": 9.974846683829501e-05,
|
| 919 |
+
"loss": 0.0351,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.038873462911543,
|
| 924 |
+
"grad_norm": 0.6237223735436187,
|
| 925 |
+
"learning_rate": 9.973083648440416e-05,
|
| 926 |
+
"loss": 0.0495,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.046806822689409,
|
| 931 |
+
"grad_norm": 0.982888186735296,
|
| 932 |
+
"learning_rate": 9.971261067537932e-05,
|
| 933 |
+
"loss": 0.0428,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.0547401824672749,
|
| 938 |
+
"grad_norm": 0.3721310946214082,
|
| 939 |
+
"learning_rate": 9.969378962944829e-05,
|
| 940 |
+
"loss": 0.0396,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.0626735422451408,
|
| 945 |
+
"grad_norm": 0.3035794519690265,
|
| 946 |
+
"learning_rate": 9.9674373571966e-05,
|
| 947 |
+
"loss": 0.0352,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.0706069020230067,
|
| 952 |
+
"grad_norm": 0.2564381074742219,
|
| 953 |
+
"learning_rate": 9.965436273541175e-05,
|
| 954 |
+
"loss": 0.034,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.0785402618008726,
|
| 959 |
+
"grad_norm": 0.2627161314780947,
|
| 960 |
+
"learning_rate": 9.963375735938651e-05,
|
| 961 |
+
"loss": 0.0328,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.0864736215787385,
|
| 966 |
+
"grad_norm": 0.23965693542315172,
|
| 967 |
+
"learning_rate": 9.961255769060997e-05,
|
| 968 |
+
"loss": 0.0318,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.0944069813566046,
|
| 973 |
+
"grad_norm": 0.24468200257324066,
|
| 974 |
+
"learning_rate": 9.95907639829176e-05,
|
| 975 |
+
"loss": 0.0312,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.1023403411344705,
|
| 980 |
+
"grad_norm": 0.2090058523545183,
|
| 981 |
+
"learning_rate": 9.95683764972577e-05,
|
| 982 |
+
"loss": 0.0301,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.1102737009123365,
|
| 987 |
+
"grad_norm": 0.2583539701995491,
|
| 988 |
+
"learning_rate": 9.954539550168816e-05,
|
| 989 |
+
"loss": 0.0301,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.1182070606902024,
|
| 994 |
+
"grad_norm": 0.22716588824206302,
|
| 995 |
+
"learning_rate": 9.952182127137327e-05,
|
| 996 |
+
"loss": 0.0293,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.1261404204680683,
|
| 1001 |
+
"grad_norm": 0.24087202301512833,
|
| 1002 |
+
"learning_rate": 9.94976540885805e-05,
|
| 1003 |
+
"loss": 0.0287,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.1340737802459342,
|
| 1008 |
+
"grad_norm": 0.23388912586015503,
|
| 1009 |
+
"learning_rate": 9.947289424267703e-05,
|
| 1010 |
+
"loss": 0.0283,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.1420071400238,
|
| 1015 |
+
"grad_norm": 0.21775302377697844,
|
| 1016 |
+
"learning_rate": 9.944754203012639e-05,
|
| 1017 |
+
"loss": 0.0273,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.149940499801666,
|
| 1022 |
+
"grad_norm": 0.2120232453240847,
|
| 1023 |
+
"learning_rate": 9.942159775448478e-05,
|
| 1024 |
+
"loss": 0.0273,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.157873859579532,
|
| 1029 |
+
"grad_norm": 0.2055361894837688,
|
| 1030 |
+
"learning_rate": 9.939506172639753e-05,
|
| 1031 |
+
"loss": 0.027,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.1658072193573978,
|
| 1036 |
+
"grad_norm": 0.22045835966574195,
|
| 1037 |
+
"learning_rate": 9.936793426359536e-05,
|
| 1038 |
+
"loss": 0.0264,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.1737405791352638,
|
| 1043 |
+
"grad_norm": 0.19936431245788608,
|
| 1044 |
+
"learning_rate": 9.934021569089058e-05,
|
| 1045 |
+
"loss": 0.0262,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.1816739389131297,
|
| 1050 |
+
"grad_norm": 0.2182877537167163,
|
| 1051 |
+
"learning_rate": 9.931190634017314e-05,
|
| 1052 |
+
"loss": 0.0256,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.1896072986909956,
|
| 1057 |
+
"grad_norm": 0.20894413268256554,
|
| 1058 |
+
"learning_rate": 9.928300655040676e-05,
|
| 1059 |
+
"loss": 0.0255,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.1975406584688615,
|
| 1064 |
+
"grad_norm": 0.217040492922683,
|
| 1065 |
+
"learning_rate": 9.925351666762479e-05,
|
| 1066 |
+
"loss": 0.0248,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.2054740182467274,
|
| 1071 |
+
"grad_norm": 0.2080798426449999,
|
| 1072 |
+
"learning_rate": 9.92234370449261e-05,
|
| 1073 |
+
"loss": 0.0247,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.2134073780245935,
|
| 1078 |
+
"grad_norm": 0.21422582180243674,
|
| 1079 |
+
"learning_rate": 9.919276804247084e-05,
|
| 1080 |
+
"loss": 0.0246,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.2213407378024592,
|
| 1085 |
+
"grad_norm": 0.22109698138341372,
|
| 1086 |
+
"learning_rate": 9.916151002747614e-05,
|
| 1087 |
+
"loss": 0.0242,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.2292740975803254,
|
| 1092 |
+
"grad_norm": 0.2191386707736148,
|
| 1093 |
+
"learning_rate": 9.912966337421169e-05,
|
| 1094 |
+
"loss": 0.0241,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.2372074573581913,
|
| 1099 |
+
"grad_norm": 0.21285746880929995,
|
| 1100 |
+
"learning_rate": 9.909722846399531e-05,
|
| 1101 |
+
"loss": 0.0235,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.2451408171360572,
|
| 1106 |
+
"grad_norm": 0.2060608693515892,
|
| 1107 |
+
"learning_rate": 9.906420568518832e-05,
|
| 1108 |
+
"loss": 0.0235,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.253074176913923,
|
| 1113 |
+
"grad_norm": 0.2165821829643641,
|
| 1114 |
+
"learning_rate": 9.903059543319096e-05,
|
| 1115 |
+
"loss": 0.0228,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.261007536691789,
|
| 1120 |
+
"grad_norm": 0.20784176983531677,
|
| 1121 |
+
"learning_rate": 9.899639811043755e-05,
|
| 1122 |
+
"loss": 0.0224,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.268940896469655,
|
| 1127 |
+
"grad_norm": 0.1941327009060369,
|
| 1128 |
+
"learning_rate": 9.896161412639178e-05,
|
| 1129 |
+
"loss": 0.0224,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.2768742562475208,
|
| 1134 |
+
"grad_norm": 0.19719426217271882,
|
| 1135 |
+
"learning_rate": 9.892624389754175e-05,
|
| 1136 |
+
"loss": 0.0218,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.2848076160253867,
|
| 1141 |
+
"grad_norm": 0.1987404294569414,
|
| 1142 |
+
"learning_rate": 9.889028784739502e-05,
|
| 1143 |
+
"loss": 0.0217,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.2927409758032526,
|
| 1148 |
+
"grad_norm": 0.36446508935174654,
|
| 1149 |
+
"learning_rate": 9.885374640647349e-05,
|
| 1150 |
+
"loss": 0.0218,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.3006743355811186,
|
| 1155 |
+
"grad_norm": 0.2153981685427444,
|
| 1156 |
+
"learning_rate": 9.881662001230827e-05,
|
| 1157 |
+
"loss": 0.0219,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.3086076953589845,
|
| 1162 |
+
"grad_norm": 0.23124643936791714,
|
| 1163 |
+
"learning_rate": 9.877890910943448e-05,
|
| 1164 |
+
"loss": 0.0215,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.3165410551368504,
|
| 1169 |
+
"grad_norm": 0.19794561936484395,
|
| 1170 |
+
"learning_rate": 9.874061414938583e-05,
|
| 1171 |
+
"loss": 0.0209,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.3244744149147163,
|
| 1176 |
+
"grad_norm": 0.1939423361167064,
|
| 1177 |
+
"learning_rate": 9.870173559068933e-05,
|
| 1178 |
+
"loss": 0.0204,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.3324077746925824,
|
| 1183 |
+
"grad_norm": 0.2157419658468051,
|
| 1184 |
+
"learning_rate": 9.86622738988597e-05,
|
| 1185 |
+
"loss": 0.02,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.3403411344704481,
|
| 1190 |
+
"grad_norm": 0.19391991216609314,
|
| 1191 |
+
"learning_rate": 9.862222954639389e-05,
|
| 1192 |
+
"loss": 0.0196,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.3482744942483142,
|
| 1197 |
+
"grad_norm": 0.19977705728669262,
|
| 1198 |
+
"learning_rate": 9.858160301276529e-05,
|
| 1199 |
+
"loss": 0.0191,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.3562078540261802,
|
| 1204 |
+
"grad_norm": 0.1658808248673058,
|
| 1205 |
+
"learning_rate": 9.854039478441817e-05,
|
| 1206 |
+
"loss": 0.0194,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.364141213804046,
|
| 1211 |
+
"grad_norm": 0.18655747400847833,
|
| 1212 |
+
"learning_rate": 9.849860535476168e-05,
|
| 1213 |
+
"loss": 0.0193,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.372074573581912,
|
| 1218 |
+
"grad_norm": 0.18591273213926962,
|
| 1219 |
+
"learning_rate": 9.845623522416407e-05,
|
| 1220 |
+
"loss": 0.0191,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.380007933359778,
|
| 1225 |
+
"grad_norm": 0.19943417251594622,
|
| 1226 |
+
"learning_rate": 9.841328489994659e-05,
|
| 1227 |
+
"loss": 0.0189,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.3879412931376438,
|
| 1232 |
+
"grad_norm": 0.20534200797212154,
|
| 1233 |
+
"learning_rate": 9.836975489637754e-05,
|
| 1234 |
+
"loss": 0.0184,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.3958746529155097,
|
| 1239 |
+
"grad_norm": 0.1911015742718259,
|
| 1240 |
+
"learning_rate": 9.832564573466598e-05,
|
| 1241 |
+
"loss": 0.0182,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.4038080126933756,
|
| 1246 |
+
"grad_norm": 0.18521458075594055,
|
| 1247 |
+
"learning_rate": 9.82809579429556e-05,
|
| 1248 |
+
"loss": 0.0178,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.4117413724712415,
|
| 1253 |
+
"grad_norm": 0.19914721587866432,
|
| 1254 |
+
"learning_rate": 9.823569205631836e-05,
|
| 1255 |
+
"loss": 0.0179,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.4196747322491075,
|
| 1260 |
+
"grad_norm": 0.19845608937674608,
|
| 1261 |
+
"learning_rate": 9.818984861674803e-05,
|
| 1262 |
+
"loss": 0.0176,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.4276080920269734,
|
| 1267 |
+
"grad_norm": 0.19674825577684443,
|
| 1268 |
+
"learning_rate": 9.814342817315374e-05,
|
| 1269 |
+
"loss": 0.0177,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.4355414518048393,
|
| 1274 |
+
"grad_norm": 0.18373733166208803,
|
| 1275 |
+
"learning_rate": 9.809643128135349e-05,
|
| 1276 |
+
"loss": 0.0171,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.4434748115827052,
|
| 1281 |
+
"grad_norm": 0.1801371927821921,
|
| 1282 |
+
"learning_rate": 9.804885850406731e-05,
|
| 1283 |
+
"loss": 0.0173,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.4514081713605713,
|
| 1288 |
+
"grad_norm": 0.17435713690708068,
|
| 1289 |
+
"learning_rate": 9.800071041091072e-05,
|
| 1290 |
+
"loss": 0.0167,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.459341531138437,
|
| 1295 |
+
"grad_norm": 0.22908445382887835,
|
| 1296 |
+
"learning_rate": 9.795198757838779e-05,
|
| 1297 |
+
"loss": 0.0168,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.4672748909163031,
|
| 1302 |
+
"grad_norm": 0.19945709262181688,
|
| 1303 |
+
"learning_rate": 9.790269058988423e-05,
|
| 1304 |
+
"loss": 0.017,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.475208250694169,
|
| 1309 |
+
"grad_norm": 0.1771428579202279,
|
| 1310 |
+
"learning_rate": 9.785282003566048e-05,
|
| 1311 |
+
"loss": 0.0169,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.483141610472035,
|
| 1316 |
+
"grad_norm": 0.16493256678410484,
|
| 1317 |
+
"learning_rate": 9.780237651284458e-05,
|
| 1318 |
+
"loss": 0.0165,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.4910749702499009,
|
| 1323 |
+
"grad_norm": 0.17934728993394863,
|
| 1324 |
+
"learning_rate": 9.775136062542509e-05,
|
| 1325 |
+
"loss": 0.0161,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.4990083300277668,
|
| 1330 |
+
"grad_norm": 0.1655943299806583,
|
| 1331 |
+
"learning_rate": 9.769977298424375e-05,
|
| 1332 |
+
"loss": 0.0157,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.5069416898056327,
|
| 1337 |
+
"grad_norm": 0.2003983514777471,
|
| 1338 |
+
"learning_rate": 9.764761420698827e-05,
|
| 1339 |
+
"loss": 0.0159,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.5148750495834986,
|
| 1344 |
+
"grad_norm": 0.17686970428136242,
|
| 1345 |
+
"learning_rate": 9.759488491818487e-05,
|
| 1346 |
+
"loss": 0.0153,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.5228084093613645,
|
| 1351 |
+
"grad_norm": 0.16545001551025132,
|
| 1352 |
+
"learning_rate": 9.754158574919082e-05,
|
| 1353 |
+
"loss": 0.0151,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.5307417691392304,
|
| 1358 |
+
"grad_norm": 0.1672222245874224,
|
| 1359 |
+
"learning_rate": 9.748771733818692e-05,
|
| 1360 |
+
"loss": 0.0149,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.5386751289170963,
|
| 1365 |
+
"grad_norm": 0.20228055159403058,
|
| 1366 |
+
"learning_rate": 9.743328033016979e-05,
|
| 1367 |
+
"loss": 0.0148,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.5466084886949623,
|
| 1372 |
+
"grad_norm": 0.1730553929947374,
|
| 1373 |
+
"learning_rate": 9.737827537694418e-05,
|
| 1374 |
+
"loss": 0.0148,
|
| 1375 |
+
"step": 1950
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.5545418484728284,
|
| 1379 |
+
"grad_norm": 0.16306713584478874,
|
| 1380 |
+
"learning_rate": 9.732270313711518e-05,
|
| 1381 |
+
"loss": 0.0145,
|
| 1382 |
+
"step": 1960
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.562475208250694,
|
| 1386 |
+
"grad_norm": 0.1613864429997223,
|
| 1387 |
+
"learning_rate": 9.726656427608032e-05,
|
| 1388 |
+
"loss": 0.0144,
|
| 1389 |
+
"step": 1970
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.5704085680285602,
|
| 1393 |
+
"grad_norm": 0.15676568016378006,
|
| 1394 |
+
"learning_rate": 9.72098594660216e-05,
|
| 1395 |
+
"loss": 0.0142,
|
| 1396 |
+
"step": 1980
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.578341927806426,
|
| 1400 |
+
"grad_norm": 0.18212353738494705,
|
| 1401 |
+
"learning_rate": 9.715258938589744e-05,
|
| 1402 |
+
"loss": 0.0138,
|
| 1403 |
+
"step": 1990
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.586275287584292,
|
| 1407 |
+
"grad_norm": 0.1724534628055386,
|
| 1408 |
+
"learning_rate": 9.709475472143453e-05,
|
| 1409 |
+
"loss": 0.0135,
|
| 1410 |
+
"step": 2000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.5942086473621577,
|
| 1414 |
+
"grad_norm": 0.2030582534730454,
|
| 1415 |
+
"learning_rate": 9.703635616511972e-05,
|
| 1416 |
+
"loss": 0.0142,
|
| 1417 |
+
"step": 2010
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.6021420071400239,
|
| 1421 |
+
"grad_norm": 0.17303461729625183,
|
| 1422 |
+
"learning_rate": 9.697739441619156e-05,
|
| 1423 |
+
"loss": 0.0136,
|
| 1424 |
+
"step": 2020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.6100753669178898,
|
| 1428 |
+
"grad_norm": 0.1670928864009479,
|
| 1429 |
+
"learning_rate": 9.69178701806321e-05,
|
| 1430 |
+
"loss": 0.0133,
|
| 1431 |
+
"step": 2030
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.6180087266957557,
|
| 1435 |
+
"grad_norm": 0.17958123638082196,
|
| 1436 |
+
"learning_rate": 9.685778417115829e-05,
|
| 1437 |
+
"loss": 0.0131,
|
| 1438 |
+
"step": 2040
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.6259420864736216,
|
| 1442 |
+
"grad_norm": 0.18217670887367746,
|
| 1443 |
+
"learning_rate": 9.679713710721356e-05,
|
| 1444 |
+
"loss": 0.0129,
|
| 1445 |
+
"step": 2050
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.6338754462514875,
|
| 1449 |
+
"grad_norm": 0.1590392893596093,
|
| 1450 |
+
"learning_rate": 9.673592971495917e-05,
|
| 1451 |
+
"loss": 0.0125,
|
| 1452 |
+
"step": 2060
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.6418088060293534,
|
| 1456 |
+
"grad_norm": 0.16546382830684306,
|
| 1457 |
+
"learning_rate": 9.667416272726542e-05,
|
| 1458 |
+
"loss": 0.0124,
|
| 1459 |
+
"step": 2070
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.6497421658072193,
|
| 1463 |
+
"grad_norm": 0.16187448696505655,
|
| 1464 |
+
"learning_rate": 9.661183688370306e-05,
|
| 1465 |
+
"loss": 0.0123,
|
| 1466 |
+
"step": 2080
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 1.6576755255850852,
|
| 1470 |
+
"grad_norm": 0.15132455192579947,
|
| 1471 |
+
"learning_rate": 9.654895293053427e-05,
|
| 1472 |
+
"loss": 0.0121,
|
| 1473 |
+
"step": 2090
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.6656088853629512,
|
| 1477 |
+
"grad_norm": 0.1570850734881935,
|
| 1478 |
+
"learning_rate": 9.648551162070385e-05,
|
| 1479 |
+
"loss": 0.0117,
|
| 1480 |
+
"step": 2100
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.6735422451408173,
|
| 1484 |
+
"grad_norm": 0.15875645484905151,
|
| 1485 |
+
"learning_rate": 9.642151371383005e-05,
|
| 1486 |
+
"loss": 0.0118,
|
| 1487 |
+
"step": 2110
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.681475604918683,
|
| 1491 |
+
"grad_norm": 0.1438981307659869,
|
| 1492 |
+
"learning_rate": 9.635695997619565e-05,
|
| 1493 |
+
"loss": 0.0112,
|
| 1494 |
+
"step": 2120
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 1.689408964696549,
|
| 1498 |
+
"grad_norm": 0.17823258729626815,
|
| 1499 |
+
"learning_rate": 9.629185118073868e-05,
|
| 1500 |
+
"loss": 0.0114,
|
| 1501 |
+
"step": 2130
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 1.6973423244744148,
|
| 1505 |
+
"grad_norm": 0.1456063020084619,
|
| 1506 |
+
"learning_rate": 9.622618810704316e-05,
|
| 1507 |
+
"loss": 0.0112,
|
| 1508 |
+
"step": 2140
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.705275684252281,
|
| 1512 |
+
"grad_norm": 0.14515257419641928,
|
| 1513 |
+
"learning_rate": 9.615997154132985e-05,
|
| 1514 |
+
"loss": 0.0111,
|
| 1515 |
+
"step": 2150
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.7132090440301466,
|
| 1519 |
+
"grad_norm": 0.14229825582089722,
|
| 1520 |
+
"learning_rate": 9.60932022764467e-05,
|
| 1521 |
+
"loss": 0.0108,
|
| 1522 |
+
"step": 2160
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.7211424038080128,
|
| 1526 |
+
"grad_norm": 0.1549735628184551,
|
| 1527 |
+
"learning_rate": 9.60258811118595e-05,
|
| 1528 |
+
"loss": 0.0107,
|
| 1529 |
+
"step": 2170
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.7290757635858787,
|
| 1533 |
+
"grad_norm": 0.16276536504847472,
|
| 1534 |
+
"learning_rate": 9.595800885364226e-05,
|
| 1535 |
+
"loss": 0.0106,
|
| 1536 |
+
"step": 2180
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.7370091233637446,
|
| 1540 |
+
"grad_norm": 0.1364292826380093,
|
| 1541 |
+
"learning_rate": 9.588958631446745e-05,
|
| 1542 |
+
"loss": 0.0106,
|
| 1543 |
+
"step": 2190
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 1.7449424831416105,
|
| 1547 |
+
"grad_norm": 0.14559551979461718,
|
| 1548 |
+
"learning_rate": 9.582061431359647e-05,
|
| 1549 |
+
"loss": 0.0103,
|
| 1550 |
+
"step": 2200
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.7528758429194764,
|
| 1554 |
+
"grad_norm": 0.1818745133582307,
|
| 1555 |
+
"learning_rate": 9.575109367686965e-05,
|
| 1556 |
+
"loss": 0.0104,
|
| 1557 |
+
"step": 2210
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.7608092026973423,
|
| 1561 |
+
"grad_norm": 0.17914978500964376,
|
| 1562 |
+
"learning_rate": 9.568102523669649e-05,
|
| 1563 |
+
"loss": 0.0099,
|
| 1564 |
+
"step": 2220
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 1.7687425624752082,
|
| 1568 |
+
"grad_norm": 0.1508452007716488,
|
| 1569 |
+
"learning_rate": 9.561040983204563e-05,
|
| 1570 |
+
"loss": 0.0103,
|
| 1571 |
+
"step": 2230
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 1.7766759222530741,
|
| 1575 |
+
"grad_norm": 0.14410972407888561,
|
| 1576 |
+
"learning_rate": 9.553924830843484e-05,
|
| 1577 |
+
"loss": 0.01,
|
| 1578 |
+
"step": 2240
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.78460928203094,
|
| 1582 |
+
"grad_norm": 0.14661639481163527,
|
| 1583 |
+
"learning_rate": 9.546754151792084e-05,
|
| 1584 |
+
"loss": 0.0096,
|
| 1585 |
+
"step": 2250
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.7925426418088062,
|
| 1589 |
+
"grad_norm": 0.13331677970140016,
|
| 1590 |
+
"learning_rate": 9.539529031908915e-05,
|
| 1591 |
+
"loss": 0.0095,
|
| 1592 |
+
"step": 2260
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.8004760015866719,
|
| 1596 |
+
"grad_norm": 0.1505833320965518,
|
| 1597 |
+
"learning_rate": 9.532249557704386e-05,
|
| 1598 |
+
"loss": 0.0093,
|
| 1599 |
+
"step": 2270
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.808409361364538,
|
| 1603 |
+
"grad_norm": 0.14264267761784652,
|
| 1604 |
+
"learning_rate": 9.52491581633971e-05,
|
| 1605 |
+
"loss": 0.0094,
|
| 1606 |
+
"step": 2280
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 1.8163427211424037,
|
| 1610 |
+
"grad_norm": 0.13122197713348774,
|
| 1611 |
+
"learning_rate": 9.517527895625878e-05,
|
| 1612 |
+
"loss": 0.0097,
|
| 1613 |
+
"step": 2290
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 1.8242760809202698,
|
| 1617 |
+
"grad_norm": 0.1395067737860133,
|
| 1618 |
+
"learning_rate": 9.510085884022597e-05,
|
| 1619 |
+
"loss": 0.009,
|
| 1620 |
+
"step": 2300
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 1.8322094406981355,
|
| 1624 |
+
"grad_norm": 0.13417802543190277,
|
| 1625 |
+
"learning_rate": 9.50258987063724e-05,
|
| 1626 |
+
"loss": 0.0091,
|
| 1627 |
+
"step": 2310
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 1.8401428004760016,
|
| 1631 |
+
"grad_norm": 0.1361399896857281,
|
| 1632 |
+
"learning_rate": 9.495039945223768e-05,
|
| 1633 |
+
"loss": 0.0091,
|
| 1634 |
+
"step": 2320
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 1.8480761602538676,
|
| 1638 |
+
"grad_norm": 0.12660004136574077,
|
| 1639 |
+
"learning_rate": 9.48743619818166e-05,
|
| 1640 |
+
"loss": 0.009,
|
| 1641 |
+
"step": 2330
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.8560095200317335,
|
| 1645 |
+
"grad_norm": 0.12929036804891955,
|
| 1646 |
+
"learning_rate": 9.47977872055484e-05,
|
| 1647 |
+
"loss": 0.0087,
|
| 1648 |
+
"step": 2340
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 1.8639428798095994,
|
| 1652 |
+
"grad_norm": 0.12111549685,
|
| 1653 |
+
"learning_rate": 9.472067604030567e-05,
|
| 1654 |
+
"loss": 0.0084,
|
| 1655 |
+
"step": 2350
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.8718762395874653,
|
| 1659 |
+
"grad_norm": 0.11410224967003375,
|
| 1660 |
+
"learning_rate": 9.464302940938359e-05,
|
| 1661 |
+
"loss": 0.0083,
|
| 1662 |
+
"step": 2360
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 1.8798095993653312,
|
| 1666 |
+
"grad_norm": 0.12355268332572798,
|
| 1667 |
+
"learning_rate": 9.45648482424887e-05,
|
| 1668 |
+
"loss": 0.0088,
|
| 1669 |
+
"step": 2370
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 1.8877429591431971,
|
| 1673 |
+
"grad_norm": 0.15793674725938558,
|
| 1674 |
+
"learning_rate": 9.448613347572786e-05,
|
| 1675 |
+
"loss": 0.0083,
|
| 1676 |
+
"step": 2380
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 1.895676318921063,
|
| 1680 |
+
"grad_norm": 0.12535858682328516,
|
| 1681 |
+
"learning_rate": 9.440688605159702e-05,
|
| 1682 |
+
"loss": 0.0084,
|
| 1683 |
+
"step": 2390
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 1.903609678698929,
|
| 1687 |
+
"grad_norm": 0.13814726939668615,
|
| 1688 |
+
"learning_rate": 9.432710691896994e-05,
|
| 1689 |
+
"loss": 0.0081,
|
| 1690 |
+
"step": 2400
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 1.911543038476795,
|
| 1694 |
+
"grad_norm": 0.3170609521576562,
|
| 1695 |
+
"learning_rate": 9.42467970330868e-05,
|
| 1696 |
+
"loss": 0.0083,
|
| 1697 |
+
"step": 2410
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 1.9194763982546608,
|
| 1701 |
+
"grad_norm": 0.1409043652874587,
|
| 1702 |
+
"learning_rate": 9.416595735554285e-05,
|
| 1703 |
+
"loss": 0.0081,
|
| 1704 |
+
"step": 2420
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 1.927409758032527,
|
| 1708 |
+
"grad_norm": 0.11628134284806954,
|
| 1709 |
+
"learning_rate": 9.408458885427673e-05,
|
| 1710 |
+
"loss": 0.008,
|
| 1711 |
+
"step": 2430
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 1.9353431178103926,
|
| 1715 |
+
"grad_norm": 0.11936220506545085,
|
| 1716 |
+
"learning_rate": 9.40026925035591e-05,
|
| 1717 |
+
"loss": 0.0077,
|
| 1718 |
+
"step": 2440
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 1.9432764775882587,
|
| 1722 |
+
"grad_norm": 0.13272648582859942,
|
| 1723 |
+
"learning_rate": 9.392026928398072e-05,
|
| 1724 |
+
"loss": 0.0078,
|
| 1725 |
+
"step": 2450
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 1.9512098373661244,
|
| 1729 |
+
"grad_norm": 0.23619264812795585,
|
| 1730 |
+
"learning_rate": 9.383732018244097e-05,
|
| 1731 |
+
"loss": 0.0078,
|
| 1732 |
+
"step": 2460
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 1.9591431971439905,
|
| 1736 |
+
"grad_norm": 0.13012160233446282,
|
| 1737 |
+
"learning_rate": 9.375384619213585e-05,
|
| 1738 |
+
"loss": 0.0078,
|
| 1739 |
+
"step": 2470
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 1.9670765569218565,
|
| 1743 |
+
"grad_norm": 0.128904077298178,
|
| 1744 |
+
"learning_rate": 9.366984831254617e-05,
|
| 1745 |
+
"loss": 0.0077,
|
| 1746 |
+
"step": 2480
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 1.9750099166997224,
|
| 1750 |
+
"grad_norm": 0.1269793451565523,
|
| 1751 |
+
"learning_rate": 9.358532754942552e-05,
|
| 1752 |
+
"loss": 0.0076,
|
| 1753 |
+
"step": 2490
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 1.9829432764775883,
|
| 1757 |
+
"grad_norm": 0.1248949950499374,
|
| 1758 |
+
"learning_rate": 9.350028491478832e-05,
|
| 1759 |
+
"loss": 0.0074,
|
| 1760 |
+
"step": 2500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 1.9908766362554542,
|
| 1764 |
+
"grad_norm": 0.11628824723897038,
|
| 1765 |
+
"learning_rate": 9.34147214268976e-05,
|
| 1766 |
+
"loss": 0.0073,
|
| 1767 |
+
"step": 2510
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 1.99880999603332,
|
| 1771 |
+
"grad_norm": 0.11767052090993746,
|
| 1772 |
+
"learning_rate": 9.33286381102529e-05,
|
| 1773 |
+
"loss": 0.007,
|
| 1774 |
+
"step": 2520
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 2.0063466878222926,
|
| 1778 |
+
"grad_norm": 0.1117903784163016,
|
| 1779 |
+
"learning_rate": 9.324203599557792e-05,
|
| 1780 |
+
"loss": 0.0059,
|
| 1781 |
+
"step": 2530
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 2.0142800476001588,
|
| 1785 |
+
"grad_norm": 0.10827884886358363,
|
| 1786 |
+
"learning_rate": 9.315491611980827e-05,
|
| 1787 |
+
"loss": 0.0057,
|
| 1788 |
+
"step": 2540
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 2.0222134073780245,
|
| 1792 |
+
"grad_norm": 0.10028381904877147,
|
| 1793 |
+
"learning_rate": 9.306727952607895e-05,
|
| 1794 |
+
"loss": 0.0052,
|
| 1795 |
+
"step": 2550
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 2.0301467671558906,
|
| 1799 |
+
"grad_norm": 0.09795247925220865,
|
| 1800 |
+
"learning_rate": 9.297912726371196e-05,
|
| 1801 |
+
"loss": 0.0052,
|
| 1802 |
+
"step": 2560
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 2.0380801269337563,
|
| 1806 |
+
"grad_norm": 0.10143825231740725,
|
| 1807 |
+
"learning_rate": 9.289046038820362e-05,
|
| 1808 |
+
"loss": 0.005,
|
| 1809 |
+
"step": 2570
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 2.0460134867116224,
|
| 1813 |
+
"grad_norm": 0.09415067512406232,
|
| 1814 |
+
"learning_rate": 9.28012799612121e-05,
|
| 1815 |
+
"loss": 0.005,
|
| 1816 |
+
"step": 2580
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 2.053946846489488,
|
| 1820 |
+
"grad_norm": 0.09590593113830713,
|
| 1821 |
+
"learning_rate": 9.271158705054453e-05,
|
| 1822 |
+
"loss": 0.005,
|
| 1823 |
+
"step": 2590
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 2.0618802062673542,
|
| 1827 |
+
"grad_norm": 0.11291094188649715,
|
| 1828 |
+
"learning_rate": 9.262138273014432e-05,
|
| 1829 |
+
"loss": 0.005,
|
| 1830 |
+
"step": 2600
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 2.0698135660452204,
|
| 1834 |
+
"grad_norm": 0.10399465747009821,
|
| 1835 |
+
"learning_rate": 9.253066808007829e-05,
|
| 1836 |
+
"loss": 0.0051,
|
| 1837 |
+
"step": 2610
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 2.077746925823086,
|
| 1841 |
+
"grad_norm": 0.09497580124640684,
|
| 1842 |
+
"learning_rate": 9.243944418652368e-05,
|
| 1843 |
+
"loss": 0.0051,
|
| 1844 |
+
"step": 2620
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 2.085680285600952,
|
| 1848 |
+
"grad_norm": 0.0993516254441346,
|
| 1849 |
+
"learning_rate": 9.234771214175526e-05,
|
| 1850 |
+
"loss": 0.0051,
|
| 1851 |
+
"step": 2630
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 2.093613645378818,
|
| 1855 |
+
"grad_norm": 0.10944500939451741,
|
| 1856 |
+
"learning_rate": 9.22554730441321e-05,
|
| 1857 |
+
"loss": 0.0051,
|
| 1858 |
+
"step": 2640
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 2.101547005156684,
|
| 1862 |
+
"grad_norm": 0.09397689887537947,
|
| 1863 |
+
"learning_rate": 9.216272799808452e-05,
|
| 1864 |
+
"loss": 0.0049,
|
| 1865 |
+
"step": 2650
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 2.1094803649345497,
|
| 1869 |
+
"grad_norm": 0.11607973538611317,
|
| 1870 |
+
"learning_rate": 9.20694781141009e-05,
|
| 1871 |
+
"loss": 0.005,
|
| 1872 |
+
"step": 2660
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 2.117413724712416,
|
| 1876 |
+
"grad_norm": 0.10689694065246308,
|
| 1877 |
+
"learning_rate": 9.197572450871424e-05,
|
| 1878 |
+
"loss": 0.0049,
|
| 1879 |
+
"step": 2670
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 2.1253470844902815,
|
| 1883 |
+
"grad_norm": 0.11797688389071613,
|
| 1884 |
+
"learning_rate": 9.188146830448895e-05,
|
| 1885 |
+
"loss": 0.0051,
|
| 1886 |
+
"step": 2680
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 2.1332804442681477,
|
| 1890 |
+
"grad_norm": 0.10007478424257744,
|
| 1891 |
+
"learning_rate": 9.17867106300073e-05,
|
| 1892 |
+
"loss": 0.0048,
|
| 1893 |
+
"step": 2690
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 2.1412138040460134,
|
| 1897 |
+
"grad_norm": 0.10670178367865263,
|
| 1898 |
+
"learning_rate": 9.169145261985593e-05,
|
| 1899 |
+
"loss": 0.0047,
|
| 1900 |
+
"step": 2700
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 2.1491471638238795,
|
| 1904 |
+
"grad_norm": 0.11105464347530923,
|
| 1905 |
+
"learning_rate": 9.159569541461232e-05,
|
| 1906 |
+
"loss": 0.0047,
|
| 1907 |
+
"step": 2710
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 2.157080523601745,
|
| 1911 |
+
"grad_norm": 0.10403957273950044,
|
| 1912 |
+
"learning_rate": 9.14994401608311e-05,
|
| 1913 |
+
"loss": 0.0047,
|
| 1914 |
+
"step": 2720
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 2.1650138833796113,
|
| 1918 |
+
"grad_norm": 0.08862815220504634,
|
| 1919 |
+
"learning_rate": 9.140268801103022e-05,
|
| 1920 |
+
"loss": 0.0048,
|
| 1921 |
+
"step": 2730
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 2.172947243157477,
|
| 1925 |
+
"grad_norm": 0.17406102479916283,
|
| 1926 |
+
"learning_rate": 9.130544012367737e-05,
|
| 1927 |
+
"loss": 0.0049,
|
| 1928 |
+
"step": 2740
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 2.180880602935343,
|
| 1932 |
+
"grad_norm": 0.9984554300367096,
|
| 1933 |
+
"learning_rate": 9.12076976631759e-05,
|
| 1934 |
+
"loss": 0.0076,
|
| 1935 |
+
"step": 2750
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 2.1888139627132093,
|
| 1939 |
+
"grad_norm": 0.1794409095899379,
|
| 1940 |
+
"learning_rate": 9.110946179985099e-05,
|
| 1941 |
+
"loss": 0.0086,
|
| 1942 |
+
"step": 2760
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 2.196747322491075,
|
| 1946 |
+
"grad_norm": 0.17099885681701577,
|
| 1947 |
+
"learning_rate": 9.10107337099356e-05,
|
| 1948 |
+
"loss": 0.0072,
|
| 1949 |
+
"step": 2770
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 2.204680682268941,
|
| 1953 |
+
"grad_norm": 0.13686932152618447,
|
| 1954 |
+
"learning_rate": 9.09115145755564e-05,
|
| 1955 |
+
"loss": 0.0063,
|
| 1956 |
+
"step": 2780
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 2.212614042046807,
|
| 1960 |
+
"grad_norm": 0.13664389983421157,
|
| 1961 |
+
"learning_rate": 9.08118055847196e-05,
|
| 1962 |
+
"loss": 0.006,
|
| 1963 |
+
"step": 2790
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 2.220547401824673,
|
| 1967 |
+
"grad_norm": 0.12880765577441727,
|
| 1968 |
+
"learning_rate": 9.071160793129675e-05,
|
| 1969 |
+
"loss": 0.0055,
|
| 1970 |
+
"step": 2800
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 2.2284807616025386,
|
| 1974 |
+
"grad_norm": 0.10465888110547632,
|
| 1975 |
+
"learning_rate": 9.061092281501044e-05,
|
| 1976 |
+
"loss": 0.0052,
|
| 1977 |
+
"step": 2810
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 2.2364141213804047,
|
| 1981 |
+
"grad_norm": 0.10776062892079107,
|
| 1982 |
+
"learning_rate": 9.050975144141988e-05,
|
| 1983 |
+
"loss": 0.0051,
|
| 1984 |
+
"step": 2820
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 2.2443474811582704,
|
| 1988 |
+
"grad_norm": 0.09673629571825465,
|
| 1989 |
+
"learning_rate": 9.040809502190654e-05,
|
| 1990 |
+
"loss": 0.0049,
|
| 1991 |
+
"step": 2830
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 2.2522808409361366,
|
| 1995 |
+
"grad_norm": 0.09846666092981253,
|
| 1996 |
+
"learning_rate": 9.030595477365967e-05,
|
| 1997 |
+
"loss": 0.0048,
|
| 1998 |
+
"step": 2840
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 2.2602142007140023,
|
| 2002 |
+
"grad_norm": 0.1694566973365221,
|
| 2003 |
+
"learning_rate": 9.020333191966155e-05,
|
| 2004 |
+
"loss": 0.0049,
|
| 2005 |
+
"step": 2850
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 2.2681475604918684,
|
| 2009 |
+
"grad_norm": 0.11789583478819116,
|
| 2010 |
+
"learning_rate": 9.010022768867306e-05,
|
| 2011 |
+
"loss": 0.0048,
|
| 2012 |
+
"step": 2860
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 2.276080920269734,
|
| 2016 |
+
"grad_norm": 0.09848282297096807,
|
| 2017 |
+
"learning_rate": 8.999664331521887e-05,
|
| 2018 |
+
"loss": 0.0046,
|
| 2019 |
+
"step": 2870
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 2.2840142800476,
|
| 2023 |
+
"grad_norm": 0.10259091930814387,
|
| 2024 |
+
"learning_rate": 8.989258003957263e-05,
|
| 2025 |
+
"loss": 0.0045,
|
| 2026 |
+
"step": 2880
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 2.291947639825466,
|
| 2030 |
+
"grad_norm": 0.11543434434269667,
|
| 2031 |
+
"learning_rate": 8.978803910774218e-05,
|
| 2032 |
+
"loss": 0.0045,
|
| 2033 |
+
"step": 2890
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 2.299880999603332,
|
| 2037 |
+
"grad_norm": 0.0936318653991697,
|
| 2038 |
+
"learning_rate": 8.968302177145461e-05,
|
| 2039 |
+
"loss": 0.0045,
|
| 2040 |
+
"step": 2900
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 2.307814359381198,
|
| 2044 |
+
"grad_norm": 0.10391048932974932,
|
| 2045 |
+
"learning_rate": 8.957752928814123e-05,
|
| 2046 |
+
"loss": 0.0045,
|
| 2047 |
+
"step": 2910
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 2.315747719159064,
|
| 2051 |
+
"grad_norm": 0.09676081245422562,
|
| 2052 |
+
"learning_rate": 8.947156292092257e-05,
|
| 2053 |
+
"loss": 0.0044,
|
| 2054 |
+
"step": 2920
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 2.3236810789369295,
|
| 2058 |
+
"grad_norm": 0.09112725418934604,
|
| 2059 |
+
"learning_rate": 8.936512393859328e-05,
|
| 2060 |
+
"loss": 0.0045,
|
| 2061 |
+
"step": 2930
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 2.3316144387147957,
|
| 2065 |
+
"grad_norm": 0.11256278250908659,
|
| 2066 |
+
"learning_rate": 8.925821361560684e-05,
|
| 2067 |
+
"loss": 0.0047,
|
| 2068 |
+
"step": 2940
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 2.339547798492662,
|
| 2072 |
+
"grad_norm": 0.09790114136420414,
|
| 2073 |
+
"learning_rate": 8.915083323206035e-05,
|
| 2074 |
+
"loss": 0.0044,
|
| 2075 |
+
"step": 2950
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 2.3474811582705275,
|
| 2079 |
+
"grad_norm": 0.09262306010758421,
|
| 2080 |
+
"learning_rate": 8.904298407367925e-05,
|
| 2081 |
+
"loss": 0.0043,
|
| 2082 |
+
"step": 2960
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 2.3554145180483936,
|
| 2086 |
+
"grad_norm": 0.08727581119763789,
|
| 2087 |
+
"learning_rate": 8.893466743180187e-05,
|
| 2088 |
+
"loss": 0.0041,
|
| 2089 |
+
"step": 2970
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 2.3633478778262593,
|
| 2093 |
+
"grad_norm": 0.08597923954803054,
|
| 2094 |
+
"learning_rate": 8.882588460336397e-05,
|
| 2095 |
+
"loss": 0.0042,
|
| 2096 |
+
"step": 2980
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 2.3712812376041255,
|
| 2100 |
+
"grad_norm": 0.09609672668945411,
|
| 2101 |
+
"learning_rate": 8.871663689088321e-05,
|
| 2102 |
+
"loss": 0.0042,
|
| 2103 |
+
"step": 2990
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 2.379214597381991,
|
| 2107 |
+
"grad_norm": 0.1367345211615606,
|
| 2108 |
+
"learning_rate": 8.860692560244361e-05,
|
| 2109 |
+
"loss": 0.0042,
|
| 2110 |
+
"step": 3000
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 2.3871479571598573,
|
| 2114 |
+
"grad_norm": 0.13457964936513436,
|
| 2115 |
+
"learning_rate": 8.849675205167982e-05,
|
| 2116 |
+
"loss": 0.0051,
|
| 2117 |
+
"step": 3010
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 2.395081316937723,
|
| 2121 |
+
"grad_norm": 0.12420549593254235,
|
| 2122 |
+
"learning_rate": 8.838611755776138e-05,
|
| 2123 |
+
"loss": 0.0046,
|
| 2124 |
+
"step": 3020
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 2.403014676715589,
|
| 2128 |
+
"grad_norm": 0.10253650256405698,
|
| 2129 |
+
"learning_rate": 8.827502344537703e-05,
|
| 2130 |
+
"loss": 0.0045,
|
| 2131 |
+
"step": 3030
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 2.410948036493455,
|
| 2135 |
+
"grad_norm": 0.08829366096110679,
|
| 2136 |
+
"learning_rate": 8.816347104471876e-05,
|
| 2137 |
+
"loss": 0.0044,
|
| 2138 |
+
"step": 3040
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 2.418881396271321,
|
| 2142 |
+
"grad_norm": 0.08614517031706824,
|
| 2143 |
+
"learning_rate": 8.805146169146585e-05,
|
| 2144 |
+
"loss": 0.0042,
|
| 2145 |
+
"step": 3050
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 2.426814756049187,
|
| 2149 |
+
"grad_norm": 0.0843132229122554,
|
| 2150 |
+
"learning_rate": 8.793899672676898e-05,
|
| 2151 |
+
"loss": 0.004,
|
| 2152 |
+
"step": 3060
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 2.4347481158270528,
|
| 2156 |
+
"grad_norm": 0.0880241777856025,
|
| 2157 |
+
"learning_rate": 8.78260774972341e-05,
|
| 2158 |
+
"loss": 0.0042,
|
| 2159 |
+
"step": 3070
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 2.4426814756049184,
|
| 2163 |
+
"grad_norm": 0.0831120676477387,
|
| 2164 |
+
"learning_rate": 8.771270535490635e-05,
|
| 2165 |
+
"loss": 0.0041,
|
| 2166 |
+
"step": 3080
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 2.4506148353827846,
|
| 2170 |
+
"grad_norm": 0.09314275488458384,
|
| 2171 |
+
"learning_rate": 8.759888165725379e-05,
|
| 2172 |
+
"loss": 0.0039,
|
| 2173 |
+
"step": 3090
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 2.4585481951606507,
|
| 2177 |
+
"grad_norm": 0.09035224703888926,
|
| 2178 |
+
"learning_rate": 8.748460776715131e-05,
|
| 2179 |
+
"loss": 0.004,
|
| 2180 |
+
"step": 3100
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 2.4664815549385164,
|
| 2184 |
+
"grad_norm": 0.08183708454038133,
|
| 2185 |
+
"learning_rate": 8.73698850528641e-05,
|
| 2186 |
+
"loss": 0.004,
|
| 2187 |
+
"step": 3110
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 2.4744149147163825,
|
| 2191 |
+
"grad_norm": 0.08278007701610446,
|
| 2192 |
+
"learning_rate": 8.725471488803141e-05,
|
| 2193 |
+
"loss": 0.0039,
|
| 2194 |
+
"step": 3120
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 2.482348274494248,
|
| 2198 |
+
"grad_norm": 0.09013010599712637,
|
| 2199 |
+
"learning_rate": 8.713909865165009e-05,
|
| 2200 |
+
"loss": 0.0039,
|
| 2201 |
+
"step": 3130
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 2.4902816342721144,
|
| 2205 |
+
"grad_norm": 0.0896002218106816,
|
| 2206 |
+
"learning_rate": 8.702303772805805e-05,
|
| 2207 |
+
"loss": 0.004,
|
| 2208 |
+
"step": 3140
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 2.49821499404998,
|
| 2212 |
+
"grad_norm": 0.0892877463855606,
|
| 2213 |
+
"learning_rate": 8.690653350691762e-05,
|
| 2214 |
+
"loss": 0.004,
|
| 2215 |
+
"step": 3150
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 2.506148353827846,
|
| 2219 |
+
"grad_norm": 0.10039406998279109,
|
| 2220 |
+
"learning_rate": 8.678958738319909e-05,
|
| 2221 |
+
"loss": 0.0038,
|
| 2222 |
+
"step": 3160
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 2.514081713605712,
|
| 2226 |
+
"grad_norm": 0.08536316767911051,
|
| 2227 |
+
"learning_rate": 8.667220075716382e-05,
|
| 2228 |
+
"loss": 0.0038,
|
| 2229 |
+
"step": 3170
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 2.522015073383578,
|
| 2233 |
+
"grad_norm": 0.07774575008129075,
|
| 2234 |
+
"learning_rate": 8.655437503434758e-05,
|
| 2235 |
+
"loss": 0.0038,
|
| 2236 |
+
"step": 3180
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 2.5299484331614437,
|
| 2240 |
+
"grad_norm": 0.08337653599038533,
|
| 2241 |
+
"learning_rate": 8.643611162554366e-05,
|
| 2242 |
+
"loss": 0.0038,
|
| 2243 |
+
"step": 3190
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 2.53788179293931,
|
| 2247 |
+
"grad_norm": 0.08902318662671499,
|
| 2248 |
+
"learning_rate": 8.631741194678607e-05,
|
| 2249 |
+
"loss": 0.0038,
|
| 2250 |
+
"step": 3200
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 2.545815152717176,
|
| 2254 |
+
"grad_norm": 0.08479123006562005,
|
| 2255 |
+
"learning_rate": 8.619827741933244e-05,
|
| 2256 |
+
"loss": 0.0037,
|
| 2257 |
+
"step": 3210
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 2.5537485124950416,
|
| 2261 |
+
"grad_norm": 0.08943099225166953,
|
| 2262 |
+
"learning_rate": 8.607870946964717e-05,
|
| 2263 |
+
"loss": 0.0038,
|
| 2264 |
+
"step": 3220
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 2.5616818722729073,
|
| 2268 |
+
"grad_norm": 0.08748118262451259,
|
| 2269 |
+
"learning_rate": 8.595870952938421e-05,
|
| 2270 |
+
"loss": 0.0038,
|
| 2271 |
+
"step": 3230
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 2.5696152320507735,
|
| 2275 |
+
"grad_norm": 0.08339662369863413,
|
| 2276 |
+
"learning_rate": 8.583827903537e-05,
|
| 2277 |
+
"loss": 0.0036,
|
| 2278 |
+
"step": 3240
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 2.5775485918286396,
|
| 2282 |
+
"grad_norm": 0.07770861567087188,
|
| 2283 |
+
"learning_rate": 8.571741942958627e-05,
|
| 2284 |
+
"loss": 0.0036,
|
| 2285 |
+
"step": 3250
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 2.5854819516065053,
|
| 2289 |
+
"grad_norm": 0.0714457769555736,
|
| 2290 |
+
"learning_rate": 8.559613215915265e-05,
|
| 2291 |
+
"loss": 0.0035,
|
| 2292 |
+
"step": 3260
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 2.5934153113843714,
|
| 2296 |
+
"grad_norm": 0.06854478917601518,
|
| 2297 |
+
"learning_rate": 8.547441867630954e-05,
|
| 2298 |
+
"loss": 0.0035,
|
| 2299 |
+
"step": 3270
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 2.601348671162237,
|
| 2303 |
+
"grad_norm": 0.09906389618096384,
|
| 2304 |
+
"learning_rate": 8.535228043840058e-05,
|
| 2305 |
+
"loss": 0.0036,
|
| 2306 |
+
"step": 3280
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 2.6092820309401032,
|
| 2310 |
+
"grad_norm": 0.08286613493780488,
|
| 2311 |
+
"learning_rate": 8.522971890785521e-05,
|
| 2312 |
+
"loss": 0.0036,
|
| 2313 |
+
"step": 3290
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 2.617215390717969,
|
| 2317 |
+
"grad_norm": 0.08091969882873308,
|
| 2318 |
+
"learning_rate": 8.510673555217125e-05,
|
| 2319 |
+
"loss": 0.0035,
|
| 2320 |
+
"step": 3300
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 2.625148750495835,
|
| 2324 |
+
"grad_norm": 0.08069042220343992,
|
| 2325 |
+
"learning_rate": 8.498333184389722e-05,
|
| 2326 |
+
"loss": 0.0036,
|
| 2327 |
+
"step": 3310
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 2.6330821102737008,
|
| 2331 |
+
"grad_norm": 0.07913790770498294,
|
| 2332 |
+
"learning_rate": 8.485950926061479e-05,
|
| 2333 |
+
"loss": 0.0036,
|
| 2334 |
+
"step": 3320
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 2.641015470051567,
|
| 2338 |
+
"grad_norm": 0.09142937985311947,
|
| 2339 |
+
"learning_rate": 8.473526928492103e-05,
|
| 2340 |
+
"loss": 0.0034,
|
| 2341 |
+
"step": 3330
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 2.6489488298294326,
|
| 2345 |
+
"grad_norm": 0.07555638243794358,
|
| 2346 |
+
"learning_rate": 8.461061340441067e-05,
|
| 2347 |
+
"loss": 0.0034,
|
| 2348 |
+
"step": 3340
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 2.6568821896072987,
|
| 2352 |
+
"grad_norm": 0.09000204886375465,
|
| 2353 |
+
"learning_rate": 8.448554311165838e-05,
|
| 2354 |
+
"loss": 0.0036,
|
| 2355 |
+
"step": 3350
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 2.664815549385165,
|
| 2359 |
+
"grad_norm": 0.08506056509395979,
|
| 2360 |
+
"learning_rate": 8.436005990420071e-05,
|
| 2361 |
+
"loss": 0.0035,
|
| 2362 |
+
"step": 3360
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 2.6727489091630305,
|
| 2366 |
+
"grad_norm": 0.08790235841873126,
|
| 2367 |
+
"learning_rate": 8.423416528451835e-05,
|
| 2368 |
+
"loss": 0.0034,
|
| 2369 |
+
"step": 3370
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 2.6806822689408962,
|
| 2373 |
+
"grad_norm": 0.09487435078751827,
|
| 2374 |
+
"learning_rate": 8.4107860760018e-05,
|
| 2375 |
+
"loss": 0.0034,
|
| 2376 |
+
"step": 3380
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 2.6886156287187624,
|
| 2380 |
+
"grad_norm": 0.07811774747886152,
|
| 2381 |
+
"learning_rate": 8.398114784301445e-05,
|
| 2382 |
+
"loss": 0.0034,
|
| 2383 |
+
"step": 3390
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 2.6965489884966285,
|
| 2387 |
+
"grad_norm": 0.0851786097650454,
|
| 2388 |
+
"learning_rate": 8.385402805071235e-05,
|
| 2389 |
+
"loss": 0.0035,
|
| 2390 |
+
"step": 3400
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 2.704482348274494,
|
| 2394 |
+
"grad_norm": 0.08323706788726688,
|
| 2395 |
+
"learning_rate": 8.372650290518809e-05,
|
| 2396 |
+
"loss": 0.0033,
|
| 2397 |
+
"step": 3410
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 2.7124157080523603,
|
| 2401 |
+
"grad_norm": 0.07906761057286403,
|
| 2402 |
+
"learning_rate": 8.359857393337161e-05,
|
| 2403 |
+
"loss": 0.0035,
|
| 2404 |
+
"step": 3420
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 2.720349067830226,
|
| 2408 |
+
"grad_norm": 0.0873769646776392,
|
| 2409 |
+
"learning_rate": 8.347024266702807e-05,
|
| 2410 |
+
"loss": 0.0034,
|
| 2411 |
+
"step": 3430
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 2.728282427608092,
|
| 2415 |
+
"grad_norm": 0.0866771190426881,
|
| 2416 |
+
"learning_rate": 8.334151064273955e-05,
|
| 2417 |
+
"loss": 0.0035,
|
| 2418 |
+
"step": 3440
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 2.736215787385958,
|
| 2422 |
+
"grad_norm": 0.08592457072874347,
|
| 2423 |
+
"learning_rate": 8.321237940188659e-05,
|
| 2424 |
+
"loss": 0.0035,
|
| 2425 |
+
"step": 3450
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 2.744149147163824,
|
| 2429 |
+
"grad_norm": 0.09029042338239049,
|
| 2430 |
+
"learning_rate": 8.308285049062983e-05,
|
| 2431 |
+
"loss": 0.0035,
|
| 2432 |
+
"step": 3460
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 2.7520825069416897,
|
| 2436 |
+
"grad_norm": 0.10069787418677155,
|
| 2437 |
+
"learning_rate": 8.29529254598914e-05,
|
| 2438 |
+
"loss": 0.0035,
|
| 2439 |
+
"step": 3470
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 2.760015866719556,
|
| 2443 |
+
"grad_norm": 0.08137290249084347,
|
| 2444 |
+
"learning_rate": 8.282260586533636e-05,
|
| 2445 |
+
"loss": 0.0034,
|
| 2446 |
+
"step": 3480
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 2.7679492264974215,
|
| 2450 |
+
"grad_norm": 0.09100832267445944,
|
| 2451 |
+
"learning_rate": 8.269189326735418e-05,
|
| 2452 |
+
"loss": 0.0032,
|
| 2453 |
+
"step": 3490
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 2.7758825862752876,
|
| 2457 |
+
"grad_norm": 0.08719461227787112,
|
| 2458 |
+
"learning_rate": 8.256078923103993e-05,
|
| 2459 |
+
"loss": 0.0031,
|
| 2460 |
+
"step": 3500
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 2.7838159460531537,
|
| 2464 |
+
"grad_norm": 0.12636511512618173,
|
| 2465 |
+
"learning_rate": 8.24292953261756e-05,
|
| 2466 |
+
"loss": 0.0034,
|
| 2467 |
+
"step": 3510
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 2.7917493058310194,
|
| 2471 |
+
"grad_norm": 0.08563728997361869,
|
| 2472 |
+
"learning_rate": 8.22974131272113e-05,
|
| 2473 |
+
"loss": 0.0038,
|
| 2474 |
+
"step": 3520
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 2.799682665608885,
|
| 2478 |
+
"grad_norm": 0.08375124152454284,
|
| 2479 |
+
"learning_rate": 8.216514421324639e-05,
|
| 2480 |
+
"loss": 0.0035,
|
| 2481 |
+
"step": 3530
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 2.8076160253867513,
|
| 2485 |
+
"grad_norm": 0.0829974589101843,
|
| 2486 |
+
"learning_rate": 8.203249016801058e-05,
|
| 2487 |
+
"loss": 0.0035,
|
| 2488 |
+
"step": 3540
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 2.8155493851646174,
|
| 2492 |
+
"grad_norm": 0.08416366583567751,
|
| 2493 |
+
"learning_rate": 8.189945257984502e-05,
|
| 2494 |
+
"loss": 0.0036,
|
| 2495 |
+
"step": 3550
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 2.823482744942483,
|
| 2499 |
+
"grad_norm": 0.07723514354970452,
|
| 2500 |
+
"learning_rate": 8.176603304168318e-05,
|
| 2501 |
+
"loss": 0.0033,
|
| 2502 |
+
"step": 3560
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 2.831416104720349,
|
| 2506 |
+
"grad_norm": 0.07850248807988759,
|
| 2507 |
+
"learning_rate": 8.163223315103184e-05,
|
| 2508 |
+
"loss": 0.0032,
|
| 2509 |
+
"step": 3570
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 2.839349464498215,
|
| 2513 |
+
"grad_norm": 0.08360156053041994,
|
| 2514 |
+
"learning_rate": 8.149805450995198e-05,
|
| 2515 |
+
"loss": 0.0033,
|
| 2516 |
+
"step": 3580
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 2.847282824276081,
|
| 2520 |
+
"grad_norm": 0.07996676728546492,
|
| 2521 |
+
"learning_rate": 8.136349872503954e-05,
|
| 2522 |
+
"loss": 0.0034,
|
| 2523 |
+
"step": 3590
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 2.8552161840539467,
|
| 2527 |
+
"grad_norm": 0.07805003653168172,
|
| 2528 |
+
"learning_rate": 8.122856740740626e-05,
|
| 2529 |
+
"loss": 0.0034,
|
| 2530 |
+
"step": 3600
|
| 2531 |
+
}
|
| 2532 |
+
],
|
| 2533 |
+
"logging_steps": 10,
|
| 2534 |
+
"max_steps": 10088,
|
| 2535 |
+
"num_input_tokens_seen": 0,
|
| 2536 |
+
"num_train_epochs": 8,
|
| 2537 |
+
"save_steps": 1800,
|
| 2538 |
+
"stateful_callbacks": {
|
| 2539 |
+
"TrainerControl": {
|
| 2540 |
+
"args": {
|
| 2541 |
+
"should_epoch_stop": false,
|
| 2542 |
+
"should_evaluate": false,
|
| 2543 |
+
"should_log": false,
|
| 2544 |
+
"should_save": true,
|
| 2545 |
+
"should_training_stop": false
|
| 2546 |
+
},
|
| 2547 |
+
"attributes": {}
|
| 2548 |
+
}
|
| 2549 |
+
},
|
| 2550 |
+
"total_flos": 259127429824512.0,
|
| 2551 |
+
"train_batch_size": 128,
|
| 2552 |
+
"trial_name": null,
|
| 2553 |
+
"trial_params": null
|
| 2554 |
+
}
|