Seal-Tools: Self-Instruct Tool Learning Dataset for Agent Tuning and Detailed Benchmark

TL;DR

Seal-Tools is a comprehensive dataset and benchmark for evaluating large language models' ability to call and compose multiple tools, including complex nested and multi-tool scenarios, advancing the development of more capable AI agents.

Contribution

The paper introduces Seal-Tools, a large-scale, self-instruct generated dataset with diverse tool instances and a new benchmark for precise evaluation of tool-calling in LLMs.

Findings

Current LLMs perform poorly on complex tool-calling tasks.

Seal-Tools includes hard and nested tool instances for rigorous testing.

The dataset enables systematic evaluation and comparison of LLM tool capabilities.

Abstract

This paper presents a new tool learning dataset Seal-Tools, which contains self-instruct API-like tools. Seal-Tools not only offers a large number of tools, but also includes instances which demonstrate the practical application of tools. Seeking to generate data on a large scale while ensuring reliability, we propose a self-instruct method to generate tools and instances, allowing precise control over the process. Moreover, our Seal-Tools contains hard instances that call multiple tools to complete the job, among which some are nested tool callings. For precise and comprehensive evaluation, we use strict format control and design three metrics from different dimensions. Therefore, Seal-Tools can serve as a new benchmark to evaluate the tool-calling ability of LLMs. Finally, we evaluate several prevalent LLMs and our finetuned model on Seal-Tools. The results show that current systems are far from perfect. The code, data and experiment results are available at https://github.com/fairyshine/Seal-Tools .