SpikeZIP-TF: Conversion is All You Need for Transformer-based SNN

TL;DR

SpikeZIP-TF presents a novel conversion method that makes Transformer-based SNNs exactly equivalent to their ANNs, achieving higher accuracy on vision and NLP tasks without accuracy loss.

Contribution

The paper introduces SpikeZIP-TF, a conversion technique ensuring exact equivalence between Transformer ANNs and SNNs, improving accuracy over existing SNN methods.

Findings

Achieves 83.82% accuracy on ImageNet

Achieves 93.79% accuracy on SST-2

Outperforms state-of-the-art Transformer-based SNNs

Abstract

Spiking neural network (SNN) has attracted great attention due to its characteristic of high efficiency and accuracy. Currently, the ANN-to-SNN conversion methods can obtain ANN on-par accuracy SNN with ultra-low latency (8 time-steps) in CNN structure on computer vision (CV) tasks. However, as Transformer-based networks have achieved prevailing precision on both CV and natural language processing (NLP), the Transformer-based SNNs are still encounting the lower accuracy w.r.t the ANN counterparts. In this work, we introduce a novel ANN-to-SNN conversion method called SpikeZIP-TF, where ANN and SNN are exactly equivalent, thus incurring no accuracy degradation. SpikeZIP-TF achieves 83.82% accuracy on CV dataset (ImageNet) and 93.79% accuracy on NLP dataset (SST-2), which are higher than SOTA Transformer-based SNNs. The code is available in GitHub: https://github.com/Intelligent-Computing-Research-Group/SpikeZIP_transformer