ToolACE: Winning the Points of LLM Function Calling
Shanghai Jiao Tong University · University of Science and Technology of China · Huawei Noah’s Ark Lab · Huawei Technologies Ltd. · Singapore Management University · Tsinghua University, Tsinghua University · The Chinese University of Hong Kong · Northwest Polytechnical University Xi'an · Huawei Noah's Ark Lab · Noah’s Ark Lab, Huawei · ICT, Chinese Academy of Sciences
PDF 由论文原始站点提供，PaperCompass 不保存论文文件。
摘要
Function calling significantly extends the application boundary of large language models (LLMs), where high-quality and diverse training data is critical for unlocking this capability. However, collecting and annotating real function-calling data is challenging, while synthetic data from existing pipelines often lack coverage and accuracy. In this paper, we present ToolACE, an automatic agentic pipeline designed to generate accurate, complex, and diverse tool-learning data, specifically tailored to the capabilities of LLMs. ToolACE leverages a novel self-evolution synthesis process to curate a comprehensive API pool of 26,507 diverse APIs. Dialogs are further generated through the interplay among multiple agents, under the guidance of a complexity evaluator. To ensure data accuracy, we implement a dual-layer verification system combining rule-based and model-based checks. We demonstrate that models trained on our synthesized data---even with only 8B parameters---achieve state-of-the-art performance, comparable to the latest GPT-4 models. Our model and a subset of the data are publicly available at https://huggingface.co/Team-ACE.