i am interested in learning more about wrnts text-to-image generation capabilities. can anyone tell me more about how this feature works and what types of images can be generated