CLUECorpus2020: A Large-scale Chinese Corpus for Pre-training Language Model

TL;DR

This paper introduces CLUECorpus2020, a large-scale Chinese corpus for pre-training language models, along with new vocabulary and models that improve efficiency and performance for Chinese NLP tasks.

Contribution

It provides a massive Chinese corpus, a compact vocabulary, and pre-trained models that enhance efficiency and achieve state-of-the-art results in Chinese language understanding.

Findings

Models trained on CLUECorpus2020 achieve excellent Chinese understanding performance.

The new 8K vocabulary reduces computational costs while maintaining accuracy.

Pre-trained models include a large version with state-of-the-art results and a tiny version for faster inference.

Abstract

In this paper, we introduce the Chinese corpus from CLUE organization, CLUECorpus2020, a large-scale corpus that can be used directly for self-supervised learning such as pre-training of a language model, or language generation. It has 100G raw corpus with 35 billion Chinese characters, which is retrieved from Common Crawl. To better understand this corpus, we conduct language understanding experiments on both small and large scale, and results show that the models trained on this corpus can achieve excellent performance on Chinese. We release a new Chinese vocabulary with a size of 8K, which is only one-third of the vocabulary size used in Chinese Bert released by Google. It saves computational cost and memory while works as good as original vocabulary. We also release both large and tiny versions of the pre-trained model on this corpus. The former achieves the state-of-the-art result, and the latter retains most precision while accelerating training and prediction speed for eight times compared to Bert-base. To facilitate future work on self-supervised learning on Chinese, we release our dataset, new vocabulary, codes, and pre-trained models on Github.