Is this your dataset? Claim it with the Hugging Face account that owns it.

speechbrain/
LoquaciousSet

Name: LoquaciousSet
Creator: speechbrain
License: https://spdx.org/licenses/CC-BY-3.0.html

Views

No views yet

	LargeScaleASR: 25,000 hours of transcribed and heterogeneous English speech recognition data for research and commercial use.

The full details are available in the paper. Made of 6 subsets:

large contains 25,000 hours of read / spontaneous and clean / noisy transcribed speech. medium contains 2,500 hours of read / spontaneous and clean / noisy transcribed speech. small contains 250 hours of read / spontaneous and clean / noisy transcribed speech. clean contains 13,000 hours of read… See the full description on the dataset page: https://huggingface.co/datasets/speechbrain/LoquaciousSet.