I made a data pipeline script to train a #MachineLearning model. The end goal is to deploy it as an API and integrated it with CI/CD but it will be (super) slow progress. You can use it as a reference and I will update it when I have free time. Next update will be add logging and deploy it as an API on Heroku.
Anyway, there are many to create a ML model and deploy it. So there is no generic rules. In this case, my method works but it not necessary work on different case so adjust it according to your needs.
To run it, check below:
To run this as API:
Unfortunately, I do not have much resources so I can deploy it on cloud. Try Heroku but I only got limited disk space.