[Request] Support text-to-video and image-to-video generation models #8527