This repository contains from-scratch PyTorch implementations for two different deep learning architectures for recommendation. There is also provided code for training and evaluating both models on the MovieLens-1M dataset.
To run, create a new directory at the root called data, containing files: ml_1m_ratings.txt
, ml_1m_users.txt
, ml_1m_movies.txt
,
which are the original MovieLens files for movie ratings, user features, and movie features.
Then, navigate into each model's diretory and run:
python3 train.py