Creates a topic model from a YouTube audio track using Latent Dirichlet Allocation (LDA) and google speech-to-text - View it on GitHub