Star-Agents: Automatic Data Optimization with LLM Agents for Instruction Tuning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhou, Hang, Tang, Yehui, Qin, Haochen, Yang, Yujie, Jin, Renren, Xiong, Deyi, Han, Kai, Wang, Yunhe
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916491962613760
author Zhou, Hang
Tang, Yehui
Qin, Haochen
Yang, Yujie
Jin, Renren
Xiong, Deyi
Han, Kai
Wang, Yunhe
author_facet Zhou, Hang
Tang, Yehui
Qin, Haochen
Yang, Yujie
Jin, Renren
Xiong, Deyi
Han, Kai
Wang, Yunhe
contents The efficacy of large language models (LLMs) on downstream tasks usually hinges on instruction tuning, which relies critically on the quality of training data. Unfortunately, collecting high-quality and diverse data is both expensive and time-consuming. To mitigate this issue, we propose a novel Star-Agents framework, which automates the enhancement of data quality across datasets through multi-agent collaboration and assessment. The framework adopts a three-pronged strategy. It initially generates diverse instruction data with multiple LLM agents through a bespoke sampling method. Subsequently, the generated data undergo a rigorous evaluation using a dual-model method that assesses both difficulty and quality. Finaly, the above process evolves in a dynamic refinement phase, where more effective LLMs are prioritized, enhancing the overall data quality. Our empirical studies, including instruction tuning experiments with models such as Pythia and LLaMA, demonstrate the effectiveness of the proposed framework. Optimized datasets have achieved substantial improvements, with an average increase of 12% and notable gains in specific metrics, such as a 40% improvement in Fermi, as evidenced by benchmarks like MT-bench, Vicuna bench, and WizardLM testset.
format Preprint
id arxiv_https___arxiv_org_abs_2411_14497
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Star-Agents: Automatic Data Optimization with LLM Agents for Instruction Tuning
Zhou, Hang
Tang, Yehui
Qin, Haochen
Yang, Yujie
Jin, Renren
Xiong, Deyi
Han, Kai
Wang, Yunhe
Computation and Language
Artificial Intelligence
The efficacy of large language models (LLMs) on downstream tasks usually hinges on instruction tuning, which relies critically on the quality of training data. Unfortunately, collecting high-quality and diverse data is both expensive and time-consuming. To mitigate this issue, we propose a novel Star-Agents framework, which automates the enhancement of data quality across datasets through multi-agent collaboration and assessment. The framework adopts a three-pronged strategy. It initially generates diverse instruction data with multiple LLM agents through a bespoke sampling method. Subsequently, the generated data undergo a rigorous evaluation using a dual-model method that assesses both difficulty and quality. Finaly, the above process evolves in a dynamic refinement phase, where more effective LLMs are prioritized, enhancing the overall data quality. Our empirical studies, including instruction tuning experiments with models such as Pythia and LLaMA, demonstrate the effectiveness of the proposed framework. Optimized datasets have achieved substantial improvements, with an average increase of 12% and notable gains in specific metrics, such as a 40% improvement in Fermi, as evidenced by benchmarks like MT-bench, Vicuna bench, and WizardLM testset.
title Star-Agents: Automatic Data Optimization with LLM Agents for Instruction Tuning
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2411.14497