AgentLite: A Lightweight Library for Building and Advancing Task-Oriented LLM Agent System

Zhiwei Liu,Weiran Yao,Jianguo Zhang,Liangwei Yang,Zuxin Liu,Juntao Tan,Prafulla K. Choubey,Tian Lan,Jason Wu,Huan Wang,Shelby Heinecke,Caiming Xiong,Silvio Savarese

2024-02-23

Abstract:The booming success of LLMs initiates rapid development in LLM agents. Though the foundation of an LLM agent is the generative model, it is critical to devise the optimal reasoning strategies and agent architectures. Accordingly, LLM agent research advances from the simple chain-of-thought prompting to more complex ReAct and Reflection reasoning strategy; agent architecture also evolves from single agent generation to multi-agent conversation, as well as multi-LLM multi-agent group chat. However, with the existing intricate frameworks and libraries, creating and evaluating new reasoning strategies and agent architectures has become a complex challenge, which hinders research investigation into LLM agents. Thus, we open-source a new AI agent library, AgentLite, which simplifies this process by offering a lightweight, user-friendly platform for innovating LLM agent reasoning, architectures, and applications with ease. AgentLite is a task-oriented framework designed to enhance the ability of agents to break down tasks and facilitate the development of multi-agent systems. Furthermore, we introduce multiple practical applications developed with AgentLite to demonstrate its convenience and flexibility. Get started now at: \url{https://github.com/SalesforceAIResearch/AgentLite}.

Artificial Intelligence,Multiagent Systems

What problem does this paper attempt to address?

The paper aims to address the challenges faced by Large Language Model (LLM) agents in research and application, particularly concerning the complexity of inference strategies and architecture design. Specifically:

1. **Simplifying the Design and Verification of New Inference Strategies and Architectures**: Existing frameworks and libraries are relatively complex, making it difficult for researchers to develop new inference strategies and architectures. Therefore, the paper proposes a lightweight and easy-to-use library—AgentLite, to simplify this process.

2. **Providing Task-Oriented Design Principles**: AgentLite focuses on task-oriented agent development, from initialization to execution, centered around specific tasks. This simplifies the design and deployment of individual agents and supports the construction of complex multi-agent systems.

3. **Supporting Multi-Level Multi-Agent Orchestration**: By managing agents (Manager Agent) to coordinate multiple individual agents (Individual Agents), a hierarchical multi-agent system structure is formed, thereby better accomplishing complex task objectives.

4. **Enhancing Practicality and Flexibility**: AgentLite not only supports various benchmark evaluations but also can be flexibly applied in real-world scenarios, promoting the development of LLM agent technology.

In summary, the goal of this paper is to accelerate the research progress of LLM agents through the AgentLite platform and to promote their widespread deployment in practical applications.