This dataset contains raw, unlabeled audio recordings primarily in the Chuvash language, collected mostly from news content (e.g., news programs, news reading, or news-related segments).
Some recordings may include Russian speech and/or code-switching.

Dataset Statistics

Total files: 255,326
Total duration: 735.8209 hours
Total seconds: 2,648,955.36

Audio Specifications

Format: Flac

Length Distribution

The dataset contains a variety of recording lengths. Below is the distribution of file durations:

License

Creative Commons Attribution Non Commercial 4.0

License and acceptable use

License: CC BY-NC 4.0 — free for research and other non-commercial use.
Commercial use: possible under a separate agreement with the SpeechCollector community — reach out on Telegram: @Michaelya.
Surveillance: please do not use this dataset to build systems for surveillance, covert monitoring, or tracking of individuals.

Project website (ru / en / cv): https://chuvash-asr.vercel.app

Downloads last month: 16

Total file size:

237 GB