Rapid Deployment of DNNs for Edge Computing via Structured Pruning at Initialization

arXiv:2404.16877v1 Announce Type: new
Abstract: Edge machine learning (ML) enables localized processing of data on devices and is underpinned by deep neural networks (DNNs). However, DNNs cannot be easily run on devices due to their substantial computing, memory and energy requirements for delivering performance that is comparable to cloud-based ML. Therefore, model compression techniques, such as pruning, have been considered. Existing pruning methods are problematic for edge ML since they: (1) Create compressed models that have limited runtime performance benefits …

abstract arxiv computing cs.ai cs.lg data deployment devices edge edge computing energy however machine machine learning memory networks neural networks performance processing pruning requirements type via