How to load local NER model for visualisation using Streamlit?


After this discussion, I decided to submit my question on this forum.

I currently discovered Streamlit as a user-friendly and accessible way of sharing Machine Learning, Data Science, and Artificial Intelligence projects. So, I decided to deploy an app in order for me to have an easier process of sharing my SpaCy NER model’s results and usage.

That is when I got stuck on a problem: I don’t know how to load my model from a different device without having the actual files that construct it and form it. More exactly, how and where should I save my model so I can access my model with spacy.load("that/path").

I suppose it should be saved somewhere in cloud, but, not integrating a lot of cloud files into my projects, I don’t consider myself secure enough to speculate. Also, this is the repo with the files for the streamlit app and I should mention that I am using SpaCy v3.0.1 (as I just saw it was released v3.1.0).

I tried using GitHub, but some of the files were over the superior limit. I know there is a way of pushing big files, but even if I succeeded in doing so, I wouldn’t know how to call the model.

Any recommendation is welcomed!