Use cached download to reduce google drive load

[aleneum]

run.py and PULSE.py attempt to download files from Google drive every time they are executed. This leads to OSError: Google Drive quota exceeded quite fast. Checking the cache directory first for synthesis.pt, mapping.pt or shape_predictor_68_face_landmarks.dat reduces server queries and also reduces initialization time.

Little remark: While automatic downloading approaches failed every time, I was able to copy the URL into my browser and download all the files manually as a workaround.

[aleneum]

Had a closer look at drive.open_url and realized caching is already there but preceeds the file name with a url hash. So I guess the continous download attempts only appear when a file cannot be downloaded automatically. I am closing this since caching is already solved in a more flexible manner and manual downloads would be detected when prefixed with the url md5hash.