This code accompanies the paper "Breaking Speech Recognizers to Imagine Lyrics," by Jon Gillick and David Bamman, presented at the workshop on Machine Learning for Creativity and Design at NeurIPS 2019.
You can play with the models using your own audio files through a demo in Google Colab here: http://bit.ly/imagining_lyrics_demo