This first version was trained primarily on close-up facial portraits. It performs best for headshots, beauty photography and upper-body portraits. Full-body consistency, clothing and specific environments were not trained in this version.