Using a simple speech recognition model in iOS with Swift