LeetCode Evaluation Dataset

该数据集是从一个知名的LeetCode Swift解决方案程序库构建而成的，包含了358个算法编程问题，旨在用于评估代码补全任务。此外，该数据集还被用来衡量Copilot for Xcode生成代码的质量，通过与基准真相的对比来评估其在AI辅助编程工具代码完成和评估任务中的性能表现。该数据集的规模涵盖了358个编程问题，专门针对代码补全以及AI辅助编程工具的性能评估任务。

This dataset is developed from a reputable LeetCode Swift solution repository, containing 358 algorithmic programming problems. It is primarily intended for evaluating code completion tasks, and also utilized to assess the code generation quality of Copilot for Xcode. Specifically, the dataset evaluates the performance of AI-assisted programming tools in code completion and related assessment tasks by comparing generated code against the ground truth. With a total of 358 programming problems, this dataset is specifically tailored for performance evaluation of code completion and AI-assisted programming tools.