About

A streamlit interface to use a news classifier accessible via an AWS Lambda function.

It consumes an environment variable LAMBDA_URL that points to an API endpoint that serves predictions from the AWS Lambda function.

Screenshot

Start UI locally

Requirements:

Docker

Make sure the lambda API service is up on running or start it. More info here
Use the docker-compose file (in the root of the project) to build and launch the streamlit service and pass the override config to set the ports and mount the local directory to the container:
```
docker compose build streamlit
docker compose -f docker-compose.yml -f docker-compose.override.yml up -d streamlit
```
Access the UI at http://localhost:8030/dashboard

Contributing to the Streamlit UI service

Navigate to the root of the project
Launch locally the streamlit and the lambda services using the dev config for docker-compose. This will mount the local directory to the container and it will allow you to change the local files and see changes in the UI straight away:
```
docker compose build streamlit lambda
docker compose -f docker-compose.yml -f docker-compose.override.yml up -d streamlit lambda
```
Make changes to the code and commit them to a new branch
Once you are happy with your changes, open a pull request