cc
Vietnamese web text extracted from Common Crawl in plaintext and Markdown.
Open dataset Updated 2026-06-18
What is cc?
The dataset comprises Vietnamese text extracted from the full Common Crawl archive and converted to plaintext and Markdown formats.
It supports large-scale Vietnamese NLP work including language-model pretraining and web-text analysis.
What you can build with cc
Train Vietnamese LLMs
Use the 10TB plaintext corpus to pretrain or continue-train language models focused on Vietnamese web language.
Build search and QA systems
Index the extracted Vietnamese pages to create domain-specific retrieval or question-answering datasets.
Study web data filtering
Analyze the provided quality filters and Vietnamese subset to benchmark Common Crawl processing pipelines.
Load cc
Python
from datasets import load_dataset
ds = load_dataset("Symato/cc")- 1pip install datasets
- 2from datasets import load_dataset
- 3ds = load_dataset('Symato/cc')
- 4Access the plaintext or markdown splits for Vietnamese content
- 5Stream or download subsets as needed for training
cc: pros & cons
Pros
- +Massive Vietnamese web corpus (~10 TB plaintext)
- +Sourced directly from Common Crawl WARC files
- +Includes both plaintext and markdown formats
- +Publicly hosted on Hugging Face
Cons
- –Extremely large size requires substantial storage and bandwidth
- –Web data remains noisy despite basic filters
- –No detailed license or usage terms provided in summary
Did you find this helpful?
Frequently asked questions
A filtered Vietnamese subset of Common Crawl containing roughly 10 TB of plaintext and markdown text.
User reviews
Verified reviews from the community shape this listing's rating.
Loading reviews…
Similar datasets
Other text & nlp options worth comparing.