PruneAug: Bridging DNN Pruning and Inference Latency on Diverse Sparse Platforms Using Automatic Layerwise Block Pruning

• Published in: IEEE transactions on computers Vol. 73; no. 11; pp. 2576 - 2589
• Main Authors: Geng, Hanfei, Liu, Yifei, Zheng, Yujie, Zhang, Li Lyna, Sun, Jingwei, Wang, Yujing, Wang, Yang, Sun, Guangzhong, Yang, Mao, Cao, Ting, Liu, Yunxin
• Format: Journal Article
• Language: English
• Published: IEEE 01.11.2024
• Subjects: Accuracy; Artificial neural networks; block pruning; Deep neural network; Hardware; Optimization; Shape; sparse kernels; Sparse matrices; Sun

Although pruning is an effective technique to reduce the number of weights in deep neural networks (DNNs), it remains challenging for the resulting sparse networks to perform low-latency inference on everyday hardware. This problem is mainly caused by the incompatibility between the unstructured spa...