Crowd Recital: Hebrew Audio for Whisper Model Training

ivrit.ai's Crowd Recital dataset is designed for training Whisper models on Hebrew audio. The dataset was last updated on 2025-11-19. It contains audio data, likely paired with text, for speech-related tasks.

Use Cases

• Fine-tuning Whisper models for Hebrew speech recognition based on the described audio data.
• Training text-to-speech models using Hebrew audio recordings mentioned in the description.
• Benchmarking automatic speech recognition performance on Hebrew audio.
• Developing multilingual speech models by incorporating Hebrew data.

Strengths

• Dataset is explicitly designed for training the Whisper model architecture.
• Released under a license that permits broad research and commercial use.
• Last updated on 2025-11-19, indicating recent maintenance.

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, file formats, and exact size are unknown, which may limit suitability assessment.

Provenance

Source

ivrit-ai

Collection Method

Likely contains crowd-sourced or collected Hebrew audio recordings.

Time Range

null

Freshness

Last updated 2025-11-19 09:49:35.

Geography

null