Beetle-Data/zh-28B
收藏官方服务：
资源简介：
Beetle-Data/zh-28B是一个中文数据集，包含预分块的文本数据，每个块由513个标记的打包序列（input_ids）组成，确保序列内无跨文档内容渗漏。数据以分片形式增量上传，上传完成后通过标记文件data/_finalized.json进行确认。
Beetle-Data/zh-28B is a Chinese-language dataset comprising pre-segmented text data. Each block consists of packed sequences of 513 tokens (input_ids), guaranteeing no cross-document content leakage within the sequences. The data is uploaded incrementally in shards, and its completion is confirmed via the file data/_finalized.json.
提供机构：
Beetle-Data
