Budget-Constrained Tool Learning with Planning (2402.15960v2)
Abstract: Despite intensive efforts devoted to tool learning, the problem of budget-constrained tool learning, which focuses on resolving user queries within a specific budget constraint, has been widely overlooked. This paper proposes a novel method for budget-constrained tool learning. Our approach involves creating a preferable plan under the budget constraint before utilizing the tools. This plan outlines the feasible tools and the maximum number of times they can be employed, offering a comprehensive overview of the tool learning process for LLMs. This allows them to allocate the budget from a broader perspective. To devise the plan without incurring significant extra costs, we suggest initially estimating the usefulness of the candidate tools based on past experience. Subsequently, we employ dynamic programming to formulate the plan. Experimental results demonstrate that our method can be integrated with various tool learning methods, significantly enhancing their effectiveness under strict budget constraints.
- Longformer: The long-document transformer. arXiv preprint arXiv:2004.05150.
- ChatCoT: Tool-augmented chain-of-thought reasoning on chat-based large language models. arXiv preprint arXiv:2305.14323.
- Training verifiers to solve math word problems. arXiv preprint arXiv:2110.14168.
- AssistGPT: A general multi-modal assistant that can plan, execute, inspect, and learn. arXiv preprint arXiv:2306.08640.
- PAL: Program-aided language models. In Proc. of ICML, volume 202, pages 10764–10799.
- CLOVA: A closed-loop visual assistant with tool usage and update. arXiv preprint arXiv:2312.10908.
- A real-world webagent with planning, long context understanding, and program synthesis. arXiv preprint arXiv:2307.12856.
- ToolkenGPT: Augmenting frozen language models with massive tools via tool embeddings. arXiv preprint arXiv:2305.11554.
- Tool documentation enables zero-shot tool-usage with large language models. arXiv preprint arXiv:2308.00675.
- AudioGPT: Understanding and generating speech, music, sound, and talking head. arXiv preprint arXiv:2304.12995.
- GeneGPT: Augmenting large language models with domain tools for improved access to biomedical information. arXiv preprint arXiv:2304.09667.
- An LLM compiler for parallel function calling. arXiv preprint arXiv:2312.04511.
- WebGLM: Towards an efficient web-enhanced question answering system with human preferences. In Proc. of SIGKDD, pages 4549–4560.
- Chameleon: Plug-and-play compositional reasoning with large language models. arXiv preprint arXiv:2304.09842.
- OpenAI. 2022. OpenAI: Introducing ChatGPT.
- ART: Automatic multi-step reasoning and tool-use for large language models. arXiv preprint arXiv:2303.09014.
- Tool learning with foundation models. arXiv preprint arXiv:2304.08354.
- ToolLLM: Facilitating large language models to master 16000+ real-world APIs. arXiv preprint arXiv:2307.16789v1.
- Toolformer: Language models can teach themselves to use tools. arXiv preprint arXiv:2302.04761.
- HuggingGPT: Solving AI tasks with chatgpt and its friends in huggingface. arXiv preprint arXiv:2303.17580.
- RestGPT: Connecting large language models with real-world applications via restful apis. arXiv preprint arXiv:2306.06624.
- Chain-of-thought prompting elicits reasoning in large language models. In Proc. of NeurIPS.
- Transformers: State-of-the-art natural language processing. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations.
- Tree of Thoughts: Deliberate problem solving with large language models. arXiv preprint arXiv:2305.10601.
- ReAct: Synergizing reasoning and acting in language models. In Proc. of ICLR.
- ToolChain*: Efficient action space navigation in large language models with a* search. arXiv preprint arXiv:2310.13227.
