How to use Menoviar28/menov90 with ESPnet:
unknown model type (must be text-to-speech or automatic-speech-recognition)
