nhagar/madlad-400_urls_clean
收藏Hugging Face2025-05-15 更新2025-08-30 收录
下载链接：
https://hf-mirror.com/datasets/nhagar/madlad-400_urls_clean
下载链接
链接失效反馈官方服务：
资源简介：
madlad-400_urls_clean数据集提供了来自allenai/MADLAD-400（清洁版本）训练记录的URLs和顶级域名。这个数据集是由Nick Hagar和Jack Bandy整理的，目的是让研究人员和实践者能够分析大型语言模型训练数据集的内容，而不需要处理数TB的原始文本数据。数据集包含两个列：URL和域名。
The madlad-400_urls_clean dataset provides URLs and top-level domains associated with training records in allenai/MADLAD-400 (clean variant). Curated by Nick Hagar and Jack Bandy, this dataset is intended to enable researchers and practitioners to analyze the contents of large language model training datasets without having to manage terabytes of raw text data. It includes two columns: URL and domain.
提供机构：
nhagar
