Decay Pruning Method: Smooth Pruning With a Self-Rectifying Procedure
This work addresses accuracy loss in neural network pruning for machine learning practitioners, offering an incremental improvement over existing methods.
The paper tackles the problem of accuracy drops in structured pruning methods by introducing the Decay Pruning Method (DPM), which uses smooth pruning and a self-rectifying mechanism to gradually reduce redundant structures, resulting in consistent performance improvements and further reductions in FLOPs when integrated with existing pruning methods.
Current structured pruning methods often result in considerable accuracy drops due to abrupt network changes and loss of information from pruned structures. To address these issues, we introduce the Decay Pruning Method (DPM), a novel smooth pruning approach with a self-rectifying mechanism. DPM consists of two key components: (i) Smooth Pruning: It converts conventional single-step pruning into multi-step smooth pruning, gradually reducing redundant structures to zero over N steps with ongoing optimization. (ii) Self-Rectifying: This procedure further enhances the aforementioned process by rectifying sub-optimal pruning based on gradient information. Our approach demonstrates strong generalizability and can be easily integrated with various existing pruning methods. We validate the effectiveness of DPM by integrating it with three popular pruning methods: OTOv2, Depgraph, and Gate Decorator. Experimental results show consistent improvements in performance compared to the original pruning methods, along with further reductions of FLOPs in most scenarios.