nick-mccormick/

tts-voices-sampler

A small sample set of audio voices, with transcriptions, from other datasets.

	Sources




	
	EmoV-DB

Link. EmoV-DB is a dataset with a forced emotional reading of an arbitrary text. Multiple emotions are supplied with the same speaker. There are only four speakers in the dataset. All have a U.S. accent. Unique non-commercial license https://github.com/numediart/EmoV-DB/blob/master/LICENSE.md

Link. Dataset which seems very… See the full description on the dataset page: https://huggingface.co/datasets/nick-mccormick/tts-voices-sampler.