11-785 Spring 2022 HW5: Unsupervised Speech Recognition