Skip to content

Latest commit

 

History

History
28 lines (21 loc) · 1.37 KB

README.md

File metadata and controls

28 lines (21 loc) · 1.37 KB
Raw

Fine Tuner

This project contains a full-stack application to help with fine tuning OpenAI's LLM's. This is project is a side project and still in development. If something seems wrong, there is a good chance it is :) Just reach out to me if you need a hand.

Installation

The easiest way to get the system running in your environment is to ensure docker-compose is installed and then run the following command:

docker-compose up

This will spin up a number of services. The code and details for each server can be found in their respective subfolders. The services, as defined in the docker-compose.yaml file are:

  • frontend: This is an nginx server that serves up a react app on port 3000
  • api: The backend Flask REST API that provides connection to the db and rabbitmq
  • db: A postres database that stores the data
  • rabbitmq: A queue to hold job information
  • consumer: This service is a sham! It should be the consumer of the queue, but it is not yet implemented. It is a placeholder for now. Instead you should manually run the "consumer.py" file from the api subfolder.

You'll also need to run the ACTUAL consumer. This is a python script that will run the actual fine tuning. To do this, you'll need to run the following script:

```shell
export OPENAI_API_KEY=<your key>
export OPENAI_ORGANIZATION=<your org id>
python ./api/workers/consumer.py