Facilitating Multi-turn Function Calling for LLMs via Compositional Instruction Tuning
This addresses the need for LLMs to handle multi-turn function calling, which is critical for compositional queries, representing an incremental improvement over single-turn approaches.
The paper tackles the problem of enabling large language models to perform multi-turn function calling for compositional real-world queries by introducing BUTTON, a method that generates synthetic compositional instruction tuning data through bottom-up instruction construction and top-down trajectory generation, resulting in a dataset of 8k data points that shows effectiveness in experiments.
Large Language Models (LLMs) have exhibited significant potential in performing diverse tasks, including the ability to call functions or use external tools to enhance their performance. While current research on function calling by LLMs primarily focuses on single-turn interactions, this paper addresses the overlooked necessity for LLMs to engage in multi-turn function calling--critical for handling compositional, real-world queries that require planning with functions but not only use functions. To facilitate this, we introduce an approach, BUTTON, which generates synthetic compositional instruction tuning data via bottom-up instruction construction and top-down trajectory generation. In the bottom-up phase, we generate simple atomic tasks based on real-world scenarios and build compositional tasks using heuristic strategies based on atomic tasks. Corresponding function definitions are then synthesized for these compositional tasks. The top-down phase features a multi-agent environment where interactions among simulated humans, assistants, and tools are utilized to gather multi-turn function calling trajectories. This approach ensures task compositionality and allows for effective function and trajectory generation by examining atomic tasks within compositional tasks. We produce a dataset BUTTONInstruct comprising 8k data points and demonstrate its effectiveness through extensive experiments across various LLMs.