Ref: Deploying ML MOdel as a Flask App on Heroku
-
Create directory and git init
-
Run python -m venv env
-
Navigate to env/Scripts and run activate
-
pip install pandas numpy scikit-learn flask gunicorn
-
pip freeze > requirements.txt
-
git add and commit requirements.txt
-
Create the pkl model file using model.py. Remember to create relevant directories (lib/models)
-
Create app.py.
-
Create main.html and store within a new 'template' directory.
-
Do flask run and make sure the web app works
-
Add README.md with documentation
-
Git add/commit these files.
-
Create github repo and push these files to the repo