Nianxiong Tan

Predicting the output structure of sparse matrix multiplication with sampled compression ratio

EPQuant: A Graph Neural Network compression approach based on product quantization