How to do face-recognition in openface using GPU?

[Hongwei Du]

Hi,

I've pulled the pre-built docker image bamos/openface and successfully done a simple face-recognition task using tools under root/openface(util/align-dlib.py, batch-represent/main.lua, demos/classifier.py train, demos/classifier.py infer).  

Now I want to try it on my GPU device (Nvidia Tesla-K80) but don't know how.

Do I have to build or run it by hand with some flags? Thanks for help.