llm-director

LLM-Director

Introduction 📖

A light-weight Python library 🐍 for deploying DAG chains of customizable Large Language Model operations. Leveraging the asyncio framework, LLM-Director allows developers to create, manage, and orchestrate actions and processes in an efficient and scalable manner. With its unique structure and integration with FastAPI, LLM-Director offes a robust solution for building applications, API endpoints, and workflow automation systems.

Quick Install⚡

To get started with LLM-Director, you can install the package using pip:

pip install llm-director

Ensure you have Python 3.7 or higher installed, as LLM-Director relies on the latest features of asyncio and FastAPI.

What You Can Build With LLM-Director 🏗️

With LLM-Director, you can create a variety of applications and systems:

• Asynchronous Workflow Systems: Design and execute complex workflows with asynchronous actions, enhancing the efficiency and responsiveness of your applications.

• API Endpoints: Easily integrate with FastAPI to expose your workflows as RESTful API endpoints, allowing for seamless interaction with front-end applications or external systems.

• Automated Data Processing Pipelines: Implement pipelines for data processing, where each step can be a distinct action, conditionally processed, and managed by the Director.

• Event-Driven Architectures: Build systems that react to specific events, trigger actions, and handle a series of responses in an orderly and scalable manner.

Documentation 📚

The LLM-Director library consists of several key components:

• Director: The central orchestrator that manages actions and listeners.
• Action: The basic unit of functionality, representing a single task or process.
• Blocks: Specialized actions such as Split, Condition, Termination, Merge, and Save, each serving a unique purpose in the workflow.
• FastAPI Integration: Methods to expose actions and workflows as API endpoints.

For detailed documentation, visit LLM-Director Documentation (Coming Soon!).

Contributing 🤝

Please Contribute! I hope to grow this fast, as I think many people will benefit from this framework.