VietVault: Large-Scale Vietnamese Text Corpus from Common Crawl (2013-2023)

VietVault is a large-scale Vietnamese language corpus curated from Common Crawl dataset dumps. It contains 80GB of raw text, cleaned and filtered for Vietnamese, sourced from dumps between 2013 and 2023. The dataset was created by author nampdn-ai and last updated on 2026-05-12.

Use Cases

• Pretrain large language models based on the 80GB of high-quality Vietnamese text.
• Train Vietnamese text classification models based on the curated corpus.
• Develop Vietnamese text generation systems based on the large-scale language data.
• Benchmark Vietnamese NLP tasks based on the cleaned and deduplicated text.
• Conduct linguistic analysis of Vietnamese based on web-scale text from 2013-2023.

Strengths

• 80GB of raw Vietnamese text provides substantial volume for model training.
• Data is described as carefully filtered and curated for high quality.
• Sourced from Common Crawl dumps spanning a 10-year period (2013-2023).

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

Common Crawl dataset dumps.

Collection Method

Filtered and curated from Common Crawl raw data.

Time Range

2013-2023

Freshness

Last updated 2026-05-12 23:27:04; freshness should be verified.