How to use

Openshift Deployment

• Go to the developer view of your Openshift cluster
• Select import from git option for an automatic deloyment
• Set git repository to https://github.com/redhat-et/foundation-models-for-documentation and the context directory to backend-app
• Click on submit and wait for s2i to build the image
• Check routes for the link to the deployed application
• Set the api keys in the deployment environment variables
• Coming soon: yaml files for deployment

Running locally

requirements.txt

Install these requirements in your environment.

Config.json

Update the config file with the path to the documents, the vector database, the embedding model name, etc.

credentials.env

The current version uses openai model that requires the OPEN_API_KEY in a credentials.env file.

Running the application

Run the app.py file in your terminal: python3 app.py

You will see the application running at 0.0.0.0:8080. You can change the hostname and the port in the app.py file.

In the UI, give a query such as "What is ROSA?" to get responses like:

Candidate documents:

1: page_content='## How does ROSA use STS?' metadata={'source': '../data/external/rosaworkshop/15-sts_expla
2:
...

Answer: 

 "ROSA is a fully-managed, turnkey application platform that allows you to focus on delivering value to your customers by building and deploying applications. Red Hat and AWS Site reliability engineering (SRE) experts manage the underlying platform so you do not have to worry about the complexity of infrastructure management. ROSA provides seamless integration with a wide range of AWS compute, database, analytics, machine learning, networking, mobile, and other services to further accelerate the building and delivering of differentiating experiences to your customers."

If the application doesn't find the vector database collection in your specified path in config.json, it will create a new collection at the same path. This operation takes several hours to finish.