High-Performance Temporal Reversible Spiking Neural Networks with $O(L)$ Training Memory and $O(1)$ Inference Cost

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Hu, JiaKui, Yao, Man, Qiu, Xuerui, Chou, Yuhong, Cai, Yuxuan, Qiao, Ning, Tian, Yonghong, XU, Bo, Li, Guoqi
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866916261173133312
author Hu, JiaKui
Yao, Man
Qiu, Xuerui
Chou, Yuhong
Cai, Yuxuan
Qiao, Ning
Tian, Yonghong
XU, Bo
Li, Guoqi
author_facet Hu, JiaKui
Yao, Man
Qiu, Xuerui
Chou, Yuhong
Cai, Yuxuan
Qiao, Ning
Tian, Yonghong
XU, Bo
Li, Guoqi
contents Multi-timestep simulation of brain-inspired Spiking Neural Networks (SNNs) boost memory requirements during training and increase inference energy cost. Current training methods cannot simultaneously solve both training and inference dilemmas. This work proposes a novel Temporal Reversible architecture for SNNs (T-RevSNN) to jointly address the training and inference challenges by altering the forward propagation of SNNs. We turn off the temporal dynamics of most spiking neurons and design multi-level temporal reversible interactions at temporal turn-on spiking neurons, resulting in a $O(L)$ training memory. Combined with the temporal reversible nature, we redesign the input encoding and network organization of SNNs to achieve $O(1)$ inference energy cost. Then, we finely adjust the internal units and residual connections of the basic SNN block to ensure the effectiveness of sparse temporal information interaction. T-RevSNN achieves excellent accuracy on ImageNet, while the memory efficiency, training time acceleration, and inference energy efficiency can be significantly improved by $8.6 \times$, $2.0 \times$, and $1.6 \times$, respectively. This work is expected to break the technical bottleneck of significantly increasing memory cost and training time for large-scale SNNs while maintaining high performance and low inference energy cost. Source code and models are available at: https://github.com/BICLab/T-RevSNN.
format Preprint
id arxiv_https___arxiv_org_abs_2405_16466
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle High-Performance Temporal Reversible Spiking Neural Networks with $O(L)$ Training Memory and $O(1)$ Inference Cost
Hu, JiaKui
Yao, Man
Qiu, Xuerui
Chou, Yuhong
Cai, Yuxuan
Qiao, Ning
Tian, Yonghong
XU, Bo
Li, Guoqi
Neural and Evolutionary Computing
Multi-timestep simulation of brain-inspired Spiking Neural Networks (SNNs) boost memory requirements during training and increase inference energy cost. Current training methods cannot simultaneously solve both training and inference dilemmas. This work proposes a novel Temporal Reversible architecture for SNNs (T-RevSNN) to jointly address the training and inference challenges by altering the forward propagation of SNNs. We turn off the temporal dynamics of most spiking neurons and design multi-level temporal reversible interactions at temporal turn-on spiking neurons, resulting in a $O(L)$ training memory. Combined with the temporal reversible nature, we redesign the input encoding and network organization of SNNs to achieve $O(1)$ inference energy cost. Then, we finely adjust the internal units and residual connections of the basic SNN block to ensure the effectiveness of sparse temporal information interaction. T-RevSNN achieves excellent accuracy on ImageNet, while the memory efficiency, training time acceleration, and inference energy efficiency can be significantly improved by $8.6 \times$, $2.0 \times$, and $1.6 \times$, respectively. This work is expected to break the technical bottleneck of significantly increasing memory cost and training time for large-scale SNNs while maintaining high performance and low inference energy cost. Source code and models are available at: https://github.com/BICLab/T-RevSNN.
title High-Performance Temporal Reversible Spiking Neural Networks with $O(L)$ Training Memory and $O(1)$ Inference Cost
topic Neural and Evolutionary Computing
url https://arxiv.org/abs/2405.16466