Large Language Models (LLMs) are key technologies driving intelligent systems to handle multiple tasks. To meet the demands of various tasks, an increasing number of LLMs-driven experts with diverse capabilities have been developed, accompanied by corresponding benchmarks to evaluate their performance. This paper proposes the Bench-CoE framework, which enables Collaboration of Experts (CoE) by effectively leveraging benchmark evaluations to achieve optimal performance across various tasks. Bench-CoE includes a set of expert models, a router for assigning tasks to corresponding experts, and a benchmark dataset for training the router. Moreover, we formulate Query-Level and Subject-Level approaches based on our framework, and analyze the merits and drawbacks of these two approaches. Finally, we conduct a series of experiments with vary data distributions on both language and multimodal tasks to validate that our proposed Bench-CoE outperforms any single model in terms of overall performance. We hope this method serves as a baseline for further research in this area. The code is available at \url{https://github.com/ZhangXJ199/Bench-CoE}.
@article{arxiv.2412.04167,
title = {Bench-CoE: a Framework for Collaboration of Experts from Benchmark},
author = {Yuanshuai Wang and Xingjian Zhang and Jinkun Zhao and Siwei Wen and Peilin Feng and Shuhao Liao and Lei Huang and Wenjun Wu},
journal= {arXiv preprint arXiv:2412.04167},
year = {2024}
}
Comments
The code is available at \url{https://github.com/ZhangXJ199/Bench-CoE}