nasa-impact/nasa-science-code-benchmark-v0.1.1
收藏Hugging Face2026-04-10 更新2026-02-07 收录
下载链接：
https://hf-mirror.com/datasets/nasa-impact/nasa-science-code-benchmark-v0.1.1
下载链接
链接失效反馈官方服务：
资源简介：
NASA代码检索基准v0.1.1是一个基于NASA GitHub仓库中7种编程语言代码的检索基准数据集。该数据集引入了层次结构，支持按语言、查询类别或NASA部门进行模型评估。数据集包含真实关系（qrels），组织为编程语言、查询类型和部门三个主要配置。README详细说明了如何根据不同评估目的加载数据集，并列出了包含的具体语言和查询类型。
The NASA Code Retrieval Benchmark v0.1.1 is a code retrieval benchmark based on code from 7 programming languages sourced from NASAs GitHub repositories. This dataset introduces a hierarchical structure, allowing model evaluation specifically by language, query category, or NASA division without data redundancy. The ground-truth relationships (qrels) are organized into three primary configurations: programming languages, query types, and divisions. The README provides detailed instructions on loading the dataset for different evaluation purposes and lists the specific languages and query types included.
提供机构：
nasa-impact
