Honk models google-speech-dataset.pt: model for discriminating keywords __silence__, __unknown__, yes, no, up, down, left, right, on, off, stop, go. Best dev accuracy: 88.5%.