TerViT: An Efficient Ternary Vision Transformer
This work addresses efficiency problems for deploying vision transformers on resource-constrained devices, representing an incremental improvement in model compression.
The paper tackles the high computational and memory costs of vision transformers on resource-constrained devices by introducing TerViT, a ternary vision transformer that ternarizes weights, achieving competitive performance such as 79% Top-1 accuracy on ImageNet with a 13.1MB model size for Swin-S.
Vision transformers (ViTs) have demonstrated great potential in various visual tasks, but suffer from expensive computational and memory cost problems when deployed on resource-constrained devices. In this paper, we introduce a ternary vision transformer (TerViT) to ternarize the weights in ViTs, which are challenged by the large loss surface gap between real-valued and ternary parameters. To address the issue, we introduce a progressive training scheme by first training 8-bit transformers and then TerViT, and achieve a better optimization than conventional methods. Furthermore, we introduce channel-wise ternarization, by partitioning each matrix to different channels, each of which is with an unique distribution and ternarization interval. We apply our methods to popular DeiT and Swin backbones, and extensive results show that we can achieve competitive performance. For example, TerViT can quantize Swin-S to 13.1MB model size while achieving above 79% Top-1 accuracy on ImageNet dataset.