Skip to content
Fan Zeng (Frank)
Back to resources

Hypa-Voices ASR Benchmark

This compact evaluation set contains verbatim reference text for recordings across widely spoken and low-resource languages.

Visit official source (opens in a new tab)

Research use

Run quick language-coverage checks on compact recognizers before scheduling a larger evaluation.

Limitations

There are few samples per language. Small groups yield noisy estimates of word error rate.

License and access

Dataset license
Apache-2.0 (opens in a new tab)
Access status
Direct access

Download the single benchmark Parquet shard, which contains FLAC audio and reference text.