Mr. TyDi is a multi-lingual benchmark dataset built on TyDi, covering eleven typologically diverse languages. - View it on GitHub