LISPから画像生成　モデル依存性が大きい？

(:generate-image (:subject (:girl (:attributes (:cat-ears true) (:expression \"impassive\")) (:clothing (:kimono \"blue\") (:skirt \"black with cherry blossoms\")) (:action \"dancing\") (:location \"riverbank\") (:companions (:cats (:count \"several\") (:action \"surrounding\"))) ) ) (:context (:time \"afternoon\") (:weather \"sunny\") (:mood \"solemn ceremony\") ) (:intent (:searching \"future\'s gate\") (:carrying \"quiet determination\") (:symbolism (:cherry-blossoms \"a dream of spring once seen\") (:expression \"hidden story within\") ) ) )