People's Speech: 30,000+ Hours of English Speech for AI Training

30,000+ hours of transcribed English speech form one of the world's largest open speech recognition corpora. The dataset was created by Srijan-Chakraborty and is licensed for academic and commercial use under CC-BY-SA and CC-BY 4.0. It was last updated on HuggingFace on 2026-01-06.

Use Cases

• Train English speech-to-text models based on the 30,000+ hours of transcribed audio.
• Benchmark automatic speech recognition (ASR) systems based on the large-scale, diverse speaker corpus.
• Develop commercial voice applications based on the permissively licensed CC-BY and CC-BY-SA content.

Strengths

• 30,000+ hours of transcribed speech provides substantial training material.
• Permissive CC-BY-SA and CC-BY 4.0 licenses allow for academic and commercial use.
• Includes a diverse set of speakers, as mentioned in the description.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and file formats are unknown, which may limit suitability assessment.
• Freshness should be verified as the last update timestamp is in the future (2026-01-06).

Provenance

Source

huggingface

Freshness

Last updated 2026-01-06 09:40:08.