Go beyond End-to-End Training: Boosting Greedy Local Learning with Context Supply

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yu, Chengting, Zhang, Fengzhao, Ma, Hanzhi, Wang, Aili, Li, Erping
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912142119141376
author Yu, Chengting
Zhang, Fengzhao
Ma, Hanzhi
Wang, Aili
Li, Erping
author_facet Yu, Chengting
Zhang, Fengzhao
Ma, Hanzhi
Wang, Aili
Li, Erping
contents Traditional end-to-end (E2E) training of deep networks necessitates storing intermediate activations for back-propagation, resulting in a large memory footprint on GPUs and restricted model parallelization. As an alternative, greedy local learning partitions the network into gradient-isolated modules and trains supervisely based on local preliminary losses, thereby providing asynchronous and parallel training methods that substantially reduce memory cost. However, empirical experiments reveal that as the number of segmentations of the gradient-isolated module increases, the performance of the local learning scheme degrades substantially, severely limiting its expansibility. To avoid this issue, we theoretically analyze the greedy local learning from the standpoint of information theory and propose a ContSup scheme, which incorporates context supply between isolated modules to compensate for information loss. Experiments on benchmark datasets (i.e. CIFAR, SVHN, STL-10) achieve SOTA results and indicate that our proposed method can significantly improve the performance of greedy local learning with minimal memory and computational overhead, allowing for the boost of the number of isolated modules. Our codes are available at https://github.com/Tab-ct/ContSup.
format Preprint
id arxiv_https___arxiv_org_abs_2312_07636
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Go beyond End-to-End Training: Boosting Greedy Local Learning with Context Supply
Yu, Chengting
Zhang, Fengzhao
Ma, Hanzhi
Wang, Aili
Li, Erping
Machine Learning
Computer Vision and Pattern Recognition
Traditional end-to-end (E2E) training of deep networks necessitates storing intermediate activations for back-propagation, resulting in a large memory footprint on GPUs and restricted model parallelization. As an alternative, greedy local learning partitions the network into gradient-isolated modules and trains supervisely based on local preliminary losses, thereby providing asynchronous and parallel training methods that substantially reduce memory cost. However, empirical experiments reveal that as the number of segmentations of the gradient-isolated module increases, the performance of the local learning scheme degrades substantially, severely limiting its expansibility. To avoid this issue, we theoretically analyze the greedy local learning from the standpoint of information theory and propose a ContSup scheme, which incorporates context supply between isolated modules to compensate for information loss. Experiments on benchmark datasets (i.e. CIFAR, SVHN, STL-10) achieve SOTA results and indicate that our proposed method can significantly improve the performance of greedy local learning with minimal memory and computational overhead, allowing for the boost of the number of isolated modules. Our codes are available at https://github.com/Tab-ct/ContSup.
title Go beyond End-to-End Training: Boosting Greedy Local Learning with Context Supply
topic Machine Learning
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2312.07636