DuReader is a benchmark dataset and model focusing on the field of machine reading comprehension, mainly used for intelligent question answering tasks.

DuReader contains 4 datasets: DuReader 2.0, DuReader Robust, DuReader Yes/No, DuReader Checklist  and DuReader-vis The main features of these datasets include:

• Real questions, real articles, real answers, and real application scenarios;
• Rich question types, including entities, numbers, opinions, etc.
• A variety of task types, including span-based tasks and classification tasks;
• Rich task challenges, including model retrieval capability, model robustness, model checklist, etc.