Tool Learning in the Wild: Empowering Language Models as Automatic Tool Agents

TL;DR

This paper introduces AutoTools, a framework that enables large language models to autonomously transform tool documentation into callable functions, integrate them into programs, and improve their ability to solve practical tasks with external tools.

Contribution

The paper presents AutoTools, a novel framework that automates tool integration for LLMs, reducing manual effort and enhancing flexibility in tool use, along with a training approach for improving open-source LLMs.

Findings

AutoTools outperforms existing methods on challenging benchmarks.

Training with synthetic data improves LLMs' tool-use capabilities.

Fine-grained training tasks enhance LLMs' understanding and programming skills.

Abstract

Augmenting large language models (LLMs) with external tools has emerged as a promising approach to extend their utility, enabling them to solve practical tasks. Previous methods manually parse tool documentation and create in-context demonstrations, transforming tools into structured formats for LLMs to use in their step-by-step reasoning. However, this manual process requires domain expertise and struggles to scale to large toolsets. Additionally, these methods rely heavily on ad-hoc inference techniques or special tokens to integrate free-form LLM generation with tool-calling actions, limiting the LLM's flexibility in handling diverse tool specifications and integrating multiple tools. In this work, we propose AutoTools, a framework that enables LLMs to automate the tool-use workflow. Specifically, the LLM automatically transforms tool documentation into callable functions, verifying syntax and runtime correctness. Then, the LLM integrates these functions into executable programs to solve practical tasks, flexibly grounding tool-use actions into its reasoning processes. Extensive experiments on existing and newly collected, more challenging benchmarks illustrate the superiority of our framework. Inspired by these promising results, we further investigate how to improve the expertise of LLMs, especially open-source LLMs with fewer parameters, within AutoTools. Thus, we propose the AutoTools-learning approach, training the LLMs with three learning tasks on 34k instances of high-quality synthetic data, including documentation understanding, relevance learning, and function programming. Fine-grained results validate the effectiveness of our overall training approach and each individual task. Our methods are an important step towards the use of LLMs for solving real-world tasks with external tools.