HumanEval
收藏arXiv2025-09-30 收录
下载链接:
https://github.com/openai/human-eval
下载链接
链接失效反馈官方服务:
资源简介:
该数据集为代码生成任务的基准,特别用于评估水印对代码生成的影响。此外,该数据集还针对pass@1得分进行了评估。所涉及的任务类型为代码生成。
This dataset is a benchmark for code generation tasks, specifically designed to evaluate the impact of watermarking on code generation. Additionally, it supports evaluations of the pass@1 score. The task type involved is code generation.



