Enabling Large Language Models (LLMs) to effectively utilize tools in multi-turn interactions is essential for building capable autonomous agents. However, acquiring diverse and realistic multi-turn tool-use data remains a significant challenge. In this work, we propose a novel text-based paradigm. We observe that textual corpora naturally contain rich, multi-step problem-solving experiences, which can serve as an untapped, scalable, and authentic data source for multi-turn tool-use tasks. Based on this insight, we introduce GEM, a data synthesis pipeline that enables the generation and extraction of multi-turn tool-use trajectories from text corpora through a four-stage process: relevance filtering, workflow & tool extraction, trajectory grounding, and complexity refinement. To reduce the computational cost, we further train a specialized Trajectory Synthesizer via supervised fine-tuning. This model distills the complex generation pipeline into an efficient, end-to-end trajectory generator. Experiments demonstrate that our GEM-32B achieve a 16.5% improvement on the BFCL V3 Multi-turn benchmark. Our models partially surpass the performance of models trained on τ - bench (Airline and Retail) in-domain data, highlighting the superior generalization capability derived from our text-based synthesis paradigm. Notably, our Trajectory Synthesizer matches the quality of the full pipeline while significantly reducing inference latency and costs.

翻译：使大型语言模型（LLMs）在多轮交互中有效利用工具，对于构建具备能力的自主智能体至关重要。然而，获取多样化且真实的多轮工具使用数据仍是一个重大挑战。在本工作中，我们提出了一种新颖的基于文本的范式。我们观察到，文本语料库天然包含丰富的多步骤问题解决经验，这些经验可作为多轮工具使用任务中一个尚未开发、可扩展且真实的数据源。基于这一洞见，我们引入了GEM，一个数据合成流程，它通过四个阶段的过程——相关性过滤、工作流与工具提取、轨迹锚定和复杂性精炼——实现从文本语料库中生成和提取多轮工具使用轨迹。为了降低计算成本，我们进一步通过监督微调训练了一个专门的轨迹合成器。该模型将复杂的生成流程提炼为一个高效的端到端轨迹生成器。实验表明，我们的GEM-32B模型在BFCL V3多轮基准测试中实现了16.5%的性能提升。我们的模型在部分指标上超越了使用τ-bench（航空和零售）领域内数据训练的模型，凸显了源自我们基于文本的合成范式的卓越泛化能力。值得注意的是，我们的轨迹合成器在显著降低推理延迟和成本的同时，其生成质量与完整流程相当。