Description

OVOS TTS plugin for Mimic2

Install

pip install ovos-tts-plugin-mimic2

Configuration

  "tts": {
    "module": "ovos-tts-plugin-mimic2",
    "ovos-tts-plugin-mimic2": {
        "voice": "kusal"
    }
  }
 

Available Voices:

• kusal - Mycroft AI official voice, hosted at https://mimic-api.mycroft.ai
• nancy - trained on Nancy Corpus by @MXGray, hosted at https://nancy.2022.us
• ljspeech - trained on LJ-Speech-Dataset by keithito, hosted at https://ljspeech.2022.us

Self Hosting

The Kusal voice model is not provided by MycroftAI and can not be self hosted

docker pull ghcr.io/openvoiceos/mimic2-nancy:dev
docker pull ghcr.io/openvoiceos/mimic2-ljspeech:dev

You can also build the containers locally

docker build -f nancy.Dockerfile -t mimic2-nancy
docker build -f ljspeech.Dockerfile -t mimic2-ljspeech

run the container

docker run --rm -p 9000:9000 mimic2-nancy

set url and voice in config, voice is used for local caching of files by ovos plugins

  "tts": {
    "module": "ovos-tts-plugin-mimic2",
    "ovos-tts-plugin-mimic2": {
        "url": "http://0.0.0.0:9000/synthesize",
        "voice": "nancy"
    }
  }