Community project: pause/speed/temperature markup + hallucination QA pipeline
Author: music-and-artsCreated Jul 23, 2026Updated Jul 23, 2026
Hi, just wanted to point to a small community project that might be relevant to #6 (adding pauses to text input).
I built (with a lot of trial-and-error and AI coding assistance, I'm not a developer myself) an extension that adds inline markup for pauses, speed and temperature directly in the input text, plus an automated pipeline that detects and regenerates hallucinated/truncated segments, with optional ASR cross-checking.
Repo: https://github.com/music-and-arts/pocket-tts-enhanced
Happy to add it under "Projects using Pocket TTS" in the README if useful, or feel free to close this if it's not a good fit — just wanted to share in case it helps someone with the same need.
Source: kyutai-labs/pocket-tts