TOOLORA / Local reading edition
Parler-TTS
Provide separate spoken text and voice descriptions to control pitch, pace, expression and recording character, or select a known speaker.
Overview
Parler-TTS separately encodes the words to speak and a style description, such as pitch, pace, expressiveness or recording distance, then generates audio. Explore random voices or name a speaker known to the model; this is not an arbitrary-reference cloning workflow. The repository provides Mini/Large inference and training code. The official Mini v1 card identifies English and Apache 2.0; language coverage and resource needs are checkpoint-specific.
- Company country / region
- United States · Company registration
Official sources and references
- Official websitehttps://github.com/huggingface/parler-tts
- Documentationhttps://huggingface.co/parler-tts/parler-tts-mini-v1
Descriptions summarize public information; consult official websites for features, pricing and regional availability.