Dataset · BAAI
Large-scale cleaned corpus of Chinese web texts
BAAI/CCI2-Data
Hundreds of millions of cleaned Chinese texts for LLM pretraining and safety.
What's inside
The second version of the CCI (Chinese Corpora Internet) corpus, comprising 100M to 1B Chinese text documents. Collected from reliable web sources and rigorously cleaned and deduplicated, it addresses the shortage of high-quality, safe Chinese data.
What you can do with it
- Pretraining and continual pretraining of Chinese LLMs
- Training Chinese content filtering models and safety systems
- Building a clean text corpus for linguistic research
- Tasks
- Text generation
- Languages
- Chinese
- Size
- 100M–1B rows
- Downloads this month
- 1.4K
- Likes
- 57
How to load it
from datasets import load_dataset
ds = load_dataset("BAAI/CCI2-Data")