Poor Man's Training on MCUs: A Memory-Efficient Quantized Back-Propagation-Free Approach

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhao, Yequan, Li, Hai, Young, Ian, Zhang, Zheng
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913575207960576
author Zhao, Yequan
Li, Hai
Young, Ian
Zhang, Zheng
author_facet Zhao, Yequan
Li, Hai
Young, Ian
Zhang, Zheng
contents Back propagation (BP) is the default solution for gradient computation in neural network training. However, implementing BP-based training on various edge devices such as FPGA, microcontrollers (MCUs), and analog computing platforms face multiple major challenges, such as the lack of hardware resources, long time-to-market, and dramatic errors in a low-precision setting. This paper presents a simple BP-free training scheme on an MCU, which makes edge training hardware design as easy as inference hardware design. We adopt a quantized zeroth-order method to estimate the gradients of quantized model parameters, which can overcome the error of a straight-through estimator in a low-precision BP scheme. We further employ a few dimension reduction methods (e.g., node perturbation, sparse training) to improve the convergence of zeroth-order training. Experiment results show that our BP-free training achieves comparable performance as BP-based training on adapting a pre-trained image classifier to various corrupted data on resource-constrained edge devices (e.g., an MCU with 1024-KB SRAM for dense full-model training, or an MCU with 256-KB SRAM for sparse training). This method is most suitable for application scenarios where memory cost and time-to-market are the major concerns, but longer latency can be tolerated.
format Preprint
id arxiv_https___arxiv_org_abs_2411_05873
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Poor Man's Training on MCUs: A Memory-Efficient Quantized Back-Propagation-Free Approach
Zhao, Yequan
Li, Hai
Young, Ian
Zhang, Zheng
Machine Learning
Computer Vision and Pattern Recognition
Distributed, Parallel, and Cluster Computing
Neural and Evolutionary Computing
I.2; C.3
Back propagation (BP) is the default solution for gradient computation in neural network training. However, implementing BP-based training on various edge devices such as FPGA, microcontrollers (MCUs), and analog computing platforms face multiple major challenges, such as the lack of hardware resources, long time-to-market, and dramatic errors in a low-precision setting. This paper presents a simple BP-free training scheme on an MCU, which makes edge training hardware design as easy as inference hardware design. We adopt a quantized zeroth-order method to estimate the gradients of quantized model parameters, which can overcome the error of a straight-through estimator in a low-precision BP scheme. We further employ a few dimension reduction methods (e.g., node perturbation, sparse training) to improve the convergence of zeroth-order training. Experiment results show that our BP-free training achieves comparable performance as BP-based training on adapting a pre-trained image classifier to various corrupted data on resource-constrained edge devices (e.g., an MCU with 1024-KB SRAM for dense full-model training, or an MCU with 256-KB SRAM for sparse training). This method is most suitable for application scenarios where memory cost and time-to-market are the major concerns, but longer latency can be tolerated.
title Poor Man's Training on MCUs: A Memory-Efficient Quantized Back-Propagation-Free Approach
topic Machine Learning
Computer Vision and Pattern Recognition
Distributed, Parallel, and Cluster Computing
Neural and Evolutionary Computing
I.2; C.3
url https://arxiv.org/abs/2411.05873