pls prepare the Python environments and read 'state.txt'

Tips: Select one of the "real_time", "split_face" or "split_eye_mouth" for the "--model" parameter.
example
split_eye_mouth:

face eye mouth