MACT: Model-Agnostic Cross-Lingual Training for Discourse Representation Structure Parsing | Read Paper on Bytez