Sparse then Prune: Toward Efficient Vision Transformers

TL;DR

This paper explores applying Sparse Regularization and Pruning to Vision Transformers to improve their efficiency and accuracy on image classification tasks, demonstrating that these methods enhance performance with reduced computational costs.

Contribution

It introduces a combined approach of Sparse Regularization and Pruning for Vision Transformers, showing improved accuracy and efficiency over traditional pruning methods.

Findings

Sparse Regularization increases accuracy by 0.12%.

Pruning with Sparse Regularization further improves accuracy, e.g., 1.764% on CIFAR-100.

The method enhances Vision Transformer performance on multiple datasets.

Abstract

The Vision Transformer architecture is a deep learning model inspired by the success of the Transformer model in Natural Language Processing. However, the self-attention mechanism, large number of parameters, and the requirement for a substantial amount of training data still make Vision Transformers computationally burdensome. In this research, we investigate the possibility of applying Sparse Regularization to Vision Transformers and the impact of Pruning, either after Sparse Regularization or without it, on the trade-off between performance and efficiency. To accomplish this, we apply Sparse Regularization and Pruning methods to the Vision Transformer architecture for image classification tasks on the CIFAR-10, CIFAR-100, and ImageNet-100 datasets. The training process for the Vision Transformer model consists of two parts: pre-training and fine-tuning. Pre-training utilizes ImageNet21K data, followed by fine-tuning for 20 epochs. The results show that when testing with CIFAR-100 and ImageNet-100 data, models with Sparse Regularization can increase accuracy by 0.12%. Furthermore, applying pruning to models with Sparse Regularization yields even better results. Specifically, it increases the average accuracy by 0.568% on CIFAR-10 data, 1.764% on CIFAR-100, and 0.256% on ImageNet-100 data compared to pruning models without Sparse Regularization. Code can be accesed here: https://github.com/yogiprsty/Sparse-ViT