As a classify user, I want to create a similar model for audio, e.g. voice classification - [ ] Create audio project type - [ ] Let user record audio for multiple classes - [ ] Let user preview the model using integrated microphone