John Thickstun

anticipation. Anticipatory Autoregressive Models

watermark. Code for watermarking language models

pytorch_musicnet. PyTorch DataSet and Jupyter demos for MusicNet

basis-separation. Implementation of the BASIS algorithm for source separation with deep generative priors

thickstun2018invariances. Experiments for Invariances and Data Augmentation for Supervised Music Transcription

thickstun2017learning. Experiments for Learning Features of Music From Scratch, ICLR 2017

mini-musicnet. The mini-MusicNet dataset

alignment-eval. Dataset and evaluation pipeline for music-to-score alignment

lean. Some experiments with the Lean proof assistant

ismir2019coupled. Experiments for Coupled Recurrent Networks for Polyphonic Music Composition

gm-hw2. Homework 2 for Generative Models

gm-hw1. Homework 1 for Generative Models

jthickstun.github.io. A beautiful, simple, clean, and responsive Jekyll theme for academics

bach-371-chorales. 371 Four-part Chorales by J.S. Bach in the Humdrum file format.

gm-hw3. Homework 3 for Generative Models

arithmetic-coding. Arithmetic coding for long-context modeling