This is the code used to produce the empyrical results reported in our paper.
Official PyTorch implementation of the paper "SOSP: Efficiently Capturing Global Correlations by Second-Order Structured Pruning"
State-of-the-art channel pruning (a.k.a. filter pruning)! This repo contains the code for ResRep: Lossless CNN Pruning via Decoupling Remembering and Forgetting.
This repository contains the official PyTorch implementation for the paper: