This repository provides a PyTorch implementation of the Graph Wavelet Neural Network (GWNN) from ICLR 2019. GWNN uses graph wavelet transform instead of graph Fourier transform to perform graph convolutions, offering computational efficiency by avoiding expensive matrix eigendecomposition. The method demonstrates improved performance over previous spectral graph CNNs on benchmark citation networks (Cora, Citeseer, Pubmed) for semi-supervised classification tasks.