STTATTS: Unified Speech-To-Text And Text-To-Speech Model