Ukrainian Speech Transcriptions from VOA Broadcasts

Encompassing transcriptions and metadata for approximately 398 hours of Ukrainian speech audio from the VOA Ukrainian dataset. It was created by Yehor and updated in February 2025 for automatic speech recognition tasks.

Use Cases

• Train an automatic speech recognition model on 398 hours of Ukrainian speech transcriptions.
• Analyze transcription metadata to understand characteristics of the VOA Ukrainian broadcast corpus.
• Fine-tune a speech model for the Ukrainian language using the provided transcriptions.

Strengths

• Contains approximately 398 hours of transcribed Ukrainian speech data.
• Specifically curated for the Ukrainian language and automatic speech recognition tasks.
• Dataset was updated in February 2025, indicating recent maintenance.

Limitations

• The dataset's size in terms of rows, columns, and file formats is unknown.
• Limited to audio from VOA Ukrainian broadcasts, which may not represent general conversational speech.
• Sample data and detailed column structure are unavailable for preview.

Provenance

Source

VOA Ukrainian dataset.

Collection Method

Transcriptions and metadata were created for the VOA audio corpus.

Freshness

Last updated on 2025-02-25.

Geography

Region is tagged as 'us', but content is in Ukrainian.