TREX: Automating LLM Fine-tuning via Agent-Driven Tree-based Exploration

2026-04-15

AI Summary

This work proposes the first end-to-end automated framework for fine-tuning large language models (LLMs), encompassing the entire lifecycle from requirement analysis to training and evaluation. The system employs a dual-module, multi-agent collaboration mechanism—comprising Researcher and Executor agents—to autonomously parse requirements, retrieve open-domain literature and data, generate training strategies, construct data recipes, and evaluate models. It further models iterative experiments as a tree-structured search space, enabling optimized exploration paths, result reuse, and high-level insight extraction. Experiments on FT-Bench, a newly introduced real-world benchmark comprising ten diverse tasks, demonstrate consistent performance gains across both general and domain-specific tasks, thereby validating the effectiveness and practicality of fully automated LLM fine-tuning.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Multiagent Systems: Multiagent Learning

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

While Large Language Models (LLMs) have empowered AI research agents to perform isolated scientific tasks, automating complex, real-world workflows, such as LLM training, remains a significant challenge. In this paper, we introduce TREX, a multi-agent system that automates the entire LLM training life-cycle. By orchestrating collaboration between two core modules-the Researcher and the Executor-the system seamlessly performs requirement analysis, open-domain literature and data research, formulation of training strategies, preparation of data recipes, and model training and evaluation. The multi-round experimental process is modeled as a search tree, enabling the system to efficiently plan exploration paths, reuse historical results, and distill high-level insights from iterative trials. To evaluate the capability of automated LLM training, we construct FT-Bench, a benchmark comprising 10 tasks derived from real-world scenarios, ranging from optimizing fundamental model capabilities to enhancing performance on domain-specific tasks. Experimental results demonstrate that the TREX agent consistently optimizes model performance on target tasks.

Problem

• LLM fine-tuning
• workflow automation
• scientific research agent
• model training lifecycle
• real-world scenarios

Innovation

• multi-agent system
• automated LLM fine-tuning
• tree-based exploration
• AI research agent
• FT-Bench