Dark mode
Uzbekistan

5.6 million documents in the Uzbek language released as open data to train artificial intelligence

Akmal Eshonxonov14 views
5.6 million documents in the Uzbek language released as open data to train artificial intelligence

The Ministry of Digital Technologies has developed a large open-text data corpus in the Uzbek language. It includes 5.6 million documents and nearly 6 billion words.

The data is presented in both Latin and Cyrillic scripts, covering texts taken from books, newspapers and archival materials, laws, web pages, and other open sources.

This corpus serves to train artificial intelligence models in the Uzbek language, improving their ability to understand Uzbek texts, provide quality answers to questions, and generate text.

To provide open access to the data, it has been uploaded to the Hugging Face platform.

Share:

Related News

5.6 million documents in the Uzbek language released as open data to train artificial intelligence | Effect.uz