Mixtraining: A Better Trade-Off Between Compute and Performance

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Zexin, Zhang, Jiancheng, Li, Yufei, Zhu, Yinglun, Liu, Cong
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913719655596032
author Li, Zexin
Zhang, Jiancheng
Li, Yufei
Zhu, Yinglun
Liu, Cong
author_facet Li, Zexin
Zhang, Jiancheng
Li, Yufei
Zhu, Yinglun
Liu, Cong
contents Incorporating self-supervised learning (SSL) before standard supervised learning (SL) has become a widely used strategy to enhance model performance, particularly in data-limited scenarios. However, this approach introduces a trade-off between computation and performance: while SSL helps with representation learning, it requires a separate, often time-consuming training phase, increasing computational overhead and limiting efficiency in resource-constrained settings. To address these challenges, we propose MixTraining, a novel framework that interleaves several SSL and SL epochs within a unified mixtraining training phase, featuring a smooth transition between two learning objectives. MixTraining enhances synergy between SSL and SL for improved accuracy and consolidates shared computation steps to reduce computation overhead. MixTraining is versatile and applicable to both single-task and multi-task learning scenarios. Extensive experiments demonstrate that MixTraining offers a superior compute-performance trade-off compared to conventional pipelines, achieving an 8.81% absolute accuracy gain (18.89% relative accuracy gain) on the TinyImageNet dataset while accelerating training by up to 1.29x with the ViT-Tiny model.
format Preprint
id arxiv_https___arxiv_org_abs_2502_19513
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Mixtraining: A Better Trade-Off Between Compute and Performance
Li, Zexin
Zhang, Jiancheng
Li, Yufei
Zhu, Yinglun
Liu, Cong
Machine Learning
Artificial Intelligence
Incorporating self-supervised learning (SSL) before standard supervised learning (SL) has become a widely used strategy to enhance model performance, particularly in data-limited scenarios. However, this approach introduces a trade-off between computation and performance: while SSL helps with representation learning, it requires a separate, often time-consuming training phase, increasing computational overhead and limiting efficiency in resource-constrained settings. To address these challenges, we propose MixTraining, a novel framework that interleaves several SSL and SL epochs within a unified mixtraining training phase, featuring a smooth transition between two learning objectives. MixTraining enhances synergy between SSL and SL for improved accuracy and consolidates shared computation steps to reduce computation overhead. MixTraining is versatile and applicable to both single-task and multi-task learning scenarios. Extensive experiments demonstrate that MixTraining offers a superior compute-performance trade-off compared to conventional pipelines, achieving an 8.81% absolute accuracy gain (18.89% relative accuracy gain) on the TinyImageNet dataset while accelerating training by up to 1.29x with the ViT-Tiny model.
title Mixtraining: A Better Trade-Off Between Compute and Performance
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2502.19513