A self-supervised learning framework for audio-visual speech - View it on GitHub