Achieving Tool Calling Functionality in LLMs Using Only Prompt Engineering Without Fine-Tuning

TL;DR

This paper introduces a prompt engineering approach that enables large language models to perform tool calling without the need for fine-tuning, achieving perfect success rates across multiple models and tasks.

Contribution

It presents a novel prompt-based method to enable tool calling in LLMs, eliminating the need for resource-intensive fine-tuning.

Findings

Achieved 100% success rate in tool calling tasks across multiple LLMs.

Demonstrated effectiveness without fine-tuning or additional training.

Applicable to various LLMs and tool calling scenarios.

Abstract

Currently, the vast majority of locally deployed open-source large language models (LLMs) and some commercial model interfaces do not support stable tool calling functionality. The existing solution involves fine-tuning LLMs, which results in significant time and computational resource consumption. This paper proposes a method that enables LLMs to achieve stable tool calling capabilities using only prompt engineering and some ingenious code design. We conducted experiments on multiple LLMs that lack tool calling capabilities across various tool calling tasks, achieving a success rate of 100%.