A Working Machine Learning Model REST API
After you create a model, if you would like other IT developers to inference from the model. API would be the best interface between you and others. Even a web developer does not understand python, who still can render predictions via API.
There are three approaches in Python you can use to create an API, using Flask, Django, and Fastapi. For beginners, Flask might be a better choice.
I tried to follow some blogs to build a machine learning API. They usually cover theory very well, but once I try to repeat the experiment, it often fails, partly because the environment might be different. Or the related packages have been updated since the post published, the original code is not able to run any more.
The following is the blog I have read.