TriviaQA Dataset

TriviaQA is a realistic text-based question answering dataset which includes 950K question-answer pairs from 662K documents collected from Wikipedia and the web. This dataset is more challenging than standard QA benchmark datasets such as Stanford Question Answering Dataset (SQuAD), as the answers for a question may not be directly obtained by span prediction and the context is very long. TriviaQA dataset consists of both human-verified and machine-generated QA subsets.

TriviaQA 是一项真实的基于文本的问答数据集，包含来自维基百科及网络上的 662,000 篇文档中的 950,000 个问题-答案对。相较于如斯坦福问答数据集（SQuAD）等标准问答基准数据集，TriviaQA 数据集更具挑战性，因为问题的答案可能无法通过直接预测文本片段获得，且上下文信息极为冗长。该数据集由经过人工验证和机器生成的问题-答案子集组成。