-
9 Feb 2024:
- Reconfigured Docker setup to have top-level Dockerfile for fly.io
- Set up auto deploy to fly.io via github actions when pushing to main
- Note that once you push to main, it will take up to a couple minutes for the changes to take affect on the public website
-
8 Feb 2024:
- Added an additional "service" of REST API build using Flask. There are only 2 simple GET endpoints
- Updated the Streamlit app to have multiple pages, show a couple other demos, and hit the above REST API if running
- Some code reorganization and tidying up
- Important: Note the change of the command to run the Streamlit app.
This is a sample app being developed for the Summer 2024 DoC in Leuven, Belgium titled Data and Software in International Government and Politics.
Currently, there are two major components:
- Streamlit App (in the
/appdirectory) - Flask REST api (in the
/apidirectory)
-
Move into the app folder with
cd app -
In your preferred python environment (or you can create a new one), install the python dependencies.
pip install -r src/requirements.txt- Run the Streamlit app locally by executing the following command (if you get an error, you may need to reload your shell/terminal for it to find the
streamlitexecutable):
streamlit run src/Home.py- More to come
- Devlop an example ML service that can perform some type of real-time inferencing. (Maybe just start off with like random number generator or something)
- Dockerize the ML App
- Automate deployment with Actions
- Find out about deploying multiple containers to Fly.io from 1 repo. Is that possible?
- Find out what the AWS equivalent of Fly.io is