metadata

license: apache-2.0

MAP-CC

An open-source Chinese pretraining dataset with a scale of 800 billion tokens, offering the NLP community high-quality Chinese pretraining data.

Usage Instructions

After downloading the parts of the dataset, you can concatenate them into a single file for each split of the dataset using the following command in a UNIX-like terminal:

cat [split].gz.part* > [split].gz

Replace [split] with the name of the dataset component you wish to merge (zh-cc, zh-baike, zh-papers, zh-books, or zh-others). After merging, decompress the .gz file to access the dataset's content.

Dataset Composition

The dataset consists of several components, each originating from different sources and serving various purposes in language modeling and processing. Below is a brief overview of each component:

Dataset Image zh-cc (Chinese Common Crawl)
Extracts from the Common Crawl project specifically filtered for Chinese content. This component is rich in diverse internet text, ranging from websites, blogs, news articles, and more.

zh-baike (Chinese Encyclopedias)
A collection of articles from various Chinese encyclopedias, similar to Wikipedia but including other encyclopedic sources as well.

zh-papers (Chinese Academic Papers)
This component consists of academic and research papers published in Chinese. It covers a wide range of disciplines and offers technical, domain-specific language.

zh-books (Chinese Books)
Comprises texts extracted from books published in Chinese. This includes literature, non-fiction, textbooks, and more.

zh-others
This category is a collection of miscellaneous texts, notably including a substantial amount of QA (Question and Answer) data, alongside a variety of other texts.

Citation

@misc{du2024chinese,
      title={Chinese Tiny LLM: Pretraining a Chinese-Centric Large Language Model}, 
      author={Xinrun Du and Zhouliang Yu and Songyang Gao and Ding Pan and Yuyang Cheng and Ziyang Ma and Ruibin Yuan and Xingwei Qu and Jiaheng Liu and Tianyu Zheng and Xinchen Luo and Guorui Zhou and Binhang Yuan and Wenhu Chen and Jie Fu and Ge Zhang},
      year={2024},
      eprint={2404.04167},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}