Feedback iteration 2

[miguelgfierro]

• [x] Make sure you are using libraries instead of having the functions in the notebooks https://github.com/Music-Recommendation-Engine-Project/Music-Recommendation-Engine/blob/main/Models/Bayesian%20Personalized%20Recommendation-2.ipynb
• [x] Create an MLOps pipeline (see other projects from teammates or follow https://github.com/miguelgfierro/project_template

[talphaidze]

We created a new folder lib inside folder Models. We created separate .py files for item-similarity class, popularity class, precision recall functions. Regarding BPR, our main recommender system, the functions that were outside the loop "with graph.as_default():", we were able to define outside the file and put it in the lib folder. However, functions for finding similarities and making recommendations, we were not able to define outside the file. This is because the session was initiated inside the loop "with graph.as_default():" and for the functions to work they need to be defined inside the loop so the session does not close. This is why, even if we try to run the functions in the same file but separate cells the BPR does not work.