AmazonScience/mxeval

MxEval是一个多语言执行评估的数据集，包含多语言编程问题，这些问题包含英文注释和文档字符串。数据集旨在评估代码生成模型的执行能力，包含多个编程任务，如MBXP、MathQA和HumanEval。数据集支持多种编程语言，包括Python、Java、JavaScript等。MxEval is a multi-lingual execution evaluation dataset containing programming problems in multiple languages, which include English comments and docstrings. The dataset is designed to evaluate the execution capabilities of code generation models, including various programming tasks such as MBXP, MathQA, and HumanEval. The dataset supports multiple programming languages, including Python, Java, JavaScript, and others.