Saved in:
Bibliographic Details
Main Authors: Qu, Yincen, Ma, Chao, Dai, Xiangying, Zhou, Hui, Wu, Yiting, Liu, Hengyue
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2411.03644
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917829796691968
author Qu, Yincen
Ma, Chao
Dai, Xiangying
Zhou, Hui
Wu, Yiting
Liu, Hengyue
author_facet Qu, Yincen
Ma, Chao
Dai, Xiangying
Zhou, Hui
Wu, Yiting
Liu, Hengyue
contents In the industry, numerous tasks are deployed online. Traditional approaches often tackle each task separately by its own network, which leads to excessive costs for developing and scaling models, especially in the context of large language models. Although multi-task methods can save costs through parameter sharing, they often struggle to outperform single-task methods in real-world applications. To tackle these challenges, we present a three-stage multi-task learning framework for large language models. It involves task filtering, followed by fine-tuning on high-resource tasks, and finally fine-tuning on all tasks. We conducted comprehensive experiments in single-task and multi-task settings. Our approach, exemplified on different benchmarks, demonstrates that it is able to achieve performance comparable to the single-task method while reducing up to 90.9\% of its overhead.
format Preprint
id arxiv_https___arxiv_org_abs_2411_03644
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Deploying Multi-task Online Server with Large Language Model
Qu, Yincen
Ma, Chao
Dai, Xiangying
Zhou, Hui
Wu, Yiting
Liu, Hengyue
Computation and Language
Artificial Intelligence
In the industry, numerous tasks are deployed online. Traditional approaches often tackle each task separately by its own network, which leads to excessive costs for developing and scaling models, especially in the context of large language models. Although multi-task methods can save costs through parameter sharing, they often struggle to outperform single-task methods in real-world applications. To tackle these challenges, we present a three-stage multi-task learning framework for large language models. It involves task filtering, followed by fine-tuning on high-resource tasks, and finally fine-tuning on all tasks. We conducted comprehensive experiments in single-task and multi-task settings. Our approach, exemplified on different benchmarks, demonstrates that it is able to achieve performance comparable to the single-task method while reducing up to 90.9\% of its overhead.
title Deploying Multi-task Online Server with Large Language Model
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2411.03644