EAPruning: Evolutionary Pruning for Vision Transformers and CNNs
This work addresses the need for efficient neural network deployment in industrial applications by offering a simple and effective pruning method that works across different network types, though it is incremental in nature.
The paper tackles the problem of structured pruning for deploying large neural networks in resource-constrained environments by proposing an evolutionary pruning approach applicable to both vision transformers and CNNs, achieving a 50% FLOPS reduction for ResNet50 and MobileNetV1 with speedups of 1.37x and 1.34x, and nearly 40% FLOPs reduction with 1.4x speedup for DeiT-Base.
Structured pruning greatly eases the deployment of large neural networks in resource-constrained environments. However, current methods either involve strong domain expertise, require extra hyperparameter tuning, or are restricted only to a specific type of network, which prevents pervasive industrial applications. In this paper, we undertake a simple and effective approach that can be easily applied to both vision transformers and convolutional neural networks. Specifically, we consider pruning as an evolution process of sub-network structures that inherit weights through reconstruction techniques. We achieve a 50% FLOPS reduction for ResNet50 and MobileNetV1, leading to 1.37x and 1.34x speedup respectively. For DeiT-Base, we reach nearly 40% FLOPs reduction and 1.4x speedup. Our code will be made available.