XQA

XQA is a cross-lingual Open-domain Question Answering dataset consisting of a training set in English and development and test sets in eight other languages. It contains question-answer pairs collected from Wikipedia, focusing on factual questions.