This version has been trained with twice the images and can produce more variation in faces, hair, and poses.
This version has been trained with twice the images and can produce more variation in faces, hair, and poses.