LGAICVNEFeb 25, 2023

A Unified Framework for Soft Threshold Pruning

arXiv:2302.13019v126 citationsh-index: 31Has Code
Originality Highly original
AI Analysis

This provides a unified framework for pruning neural networks, addressing a theoretical gap for researchers and practitioners, though it is incremental in building on existing soft threshold methods.

The paper tackles the lack of a theoretical foundation for soft threshold pruning by reformulating it as an implicit optimization problem using ISTA, unifying previous methods as variations of tuning an L1-regularization term, and derives an optimal threshold scheduler that achieves state-of-the-art performance, e.g., on ResNet-50 and MobileNet-V1 on ImageNet.

Soft threshold pruning is among the cutting-edge pruning methods with state-of-the-art performance. However, previous methods either perform aimless searching on the threshold scheduler or simply set the threshold trainable, lacking theoretical explanation from a unified perspective. In this work, we reformulate soft threshold pruning as an implicit optimization problem solved using the Iterative Shrinkage-Thresholding Algorithm (ISTA), a classic method from the fields of sparse recovery and compressed sensing. Under this theoretical framework, all threshold tuning strategies proposed in previous studies of soft threshold pruning are concluded as different styles of tuning $L_1$-regularization term. We further derive an optimal threshold scheduler through an in-depth study of threshold scheduling based on our framework. This scheduler keeps $L_1$-regularization coefficient stable, implying a time-invariant objective function from the perspective of optimization. In principle, the derived pruning algorithm could sparsify any mathematical model trained via SGD. We conduct extensive experiments and verify its state-of-the-art performance on both Artificial Neural Networks (ResNet-50 and MobileNet-V1) and Spiking Neural Networks (SEW ResNet-18) on ImageNet datasets. On the basis of this framework, we derive a family of pruning methods, including sparsify-during-training, early pruning, and pruning at initialization. The code is available at https://github.com/Yanqi-Chen/LATS.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes