As far as I know they have not released those phonetic recordings. But even without those recordings it might be possible to use those deepfake voice fingerprinting systems to build an STT engine from sound clips from the show.