Speech models for MemeD's Private voice triggers (from the Vosk project). Nothing to see here: memed.stream