CodeQA

该数据集由函数、问题和简短答案组成，这些内容是通过语法解析从代码注释语料库中生成的。它包含了“什么”、“哪里”、“何时”、“如何”、“为什么”等类型的问题以及是非问题，并分为Java和Python两个子集。该数据集的任务是对源代码进行问题回答。

This dataset comprises functions, questions and short answers, which are generated from a code comment corpus through syntactic parsing. It covers questions of types including what, where, when, how, why, as well as yes-no questions, and is split into two subsets for Java and Python. The core task of this dataset is source code-based question answering.