Introduction

Welcome to the Agent Framework documentation. This framework integrates advanced Large Language Models (LLMs) such as GPT-3.5, LLaMA, and OpenChat, providing a versatile environment for developing applications with sophisticated natural language processing, reasoning, and communication capabilities.

Purpose

The framework is designed to facilitate the development of applications that leverage the power of LLMs. Its primary goal is to provide an easy-to-use interface for switching between various prompting methods and supporting fair testing environments across different LLMs.

Key Features

• Integration with Multiple LLMs: Agent supports various models like GPT-3.5, LLaMA, and OpenChat, allowing for flexible application development.

• Versatile Prompting Methods: The framework enables easy switching between different prompting methods, catering to a wide range of use cases.

• Comprehensive Configuration Management: Using Hydra, Agent offers a robust configuration system, simplifying the management of different settings and environments.

• Advanced Memory Management: The framework includes sophisticated memory handling, enabling more complex reasoning and decision-making capabilities.

• Tool Integration: The framework includes various tools for tasks such as code interpretation, providing a comprehensive environment for application development.

• Several Testing Environments: Agent supports multiple testing environments, with the ability to easily extend to more environments.

Getting Started

To begin using the Agent Framework, please refer to the Installation Guide guide. This will provide you with the necessary steps to set up the framework and start developing your applications.

Community and Contributions

The Agent Framework is an open-source project, and we welcome contributions from the community. Whether it’s adding new features, fixing bugs, or improving documentation, your contributions are invaluable to the project’s growth and success.

Next Steps

After setting up the framework, you might want to:

• Explore what LLMs, Tasks, and Methods mean within Agent.

• Understand the building blocks of the underlying framework, such as Memory, Flows, and Intrinsic Functions.

• Explore the notebook tutorials in the GitLab repository:

  • Running Agent

  • Adding your own LLM

  • Creating your own task

  • Creating your own method

• Leverage the documentation and GitLab customization tutorials to start your own project.