This is an experimental model attempting to blend two very different sets of (ugly) training session epoch sample images into a single style.