parameterlab/scaling_mia_the_pile_00_Pile-CC
收藏Hugging Face2024-09-24 更新2025-02-15 收录
下载链接：
https://hf-mirror.com/datasets/parameterlab/scaling_mia_the_pile_00_Pile-CC
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含两个主要字段：文本内容(text)和元数据(meta)，其中元数据可能包含堆集集名称等信息。数据集分为训练集、验证集和测试集，分别包含1,748,721、52,792和52,790个示例。总数据大小为8,230,036,444字节，下载大小为4,905,482,407字节。
The dataset includes two main fields: text content (text) and metadata (meta), which may contain information such as pile set names. The dataset is split into training, validation, and test sets, containing 1,748,721, 52,792, and 52,790 examples respectively. The total dataset size is 8,230,036,444 bytes, with a download size of 4,905,482,407 bytes.
提供机构：
parameterlab
