ai_agent

This is meant to be a template for creating multi-agent workloads in python leveraging langchain/graph, fastapi, openai, and open-webui as a chat interface. The intent is to get a working scaffolding setup that I can build upon to make personal and professional assistants. The API should be openai compatible to support any client/front end.

Requirements

• Docker installed on your system.
• LM Studio installed and running.
  • Download and run the model of your choice, updating the model reference within main.py

Building the Container

1. Setup project

   export PROJECT=ai_agent

2. Clone the repository into folder named based on variable $PROJECT

   git clone https://github.com/jeffbeagley/<repo>.git $PROJECT
   cd $PROJECT

3. Build the Docker image:

   docker build -t $PROJECT .

Running the Project

Use the provided run.sh script to start the application.

./run.sh 

Integrate into your interface/client that supports openai spec.