Since this is just a test using BentoML I created a naive Machine Learning model with Iris dataset. It consists of a classifier to predict one of the 4 corresponding labels.
Then, I used BentML features to persist this model and automatically set a version tag to it.
Finally, FastAPI comes into play by loading the saved model and making predictions through an API using the same neat features provided by BentoML.

1. Execute the notebook until the end, the goal here is having a saved model into your user directory.
2. Install tox with ´pip install tox´.
3. Browse to ´fastapi´ directory and run ´tox -e python3.11-run´.
4. Go to http://127.0.0.1:3000/