Please note that this implementation of IGMC is not completely time-efficient.
It has only been correctly been trained and tested on the MovieLens100K dataset, and scalability is not ensured. After all, I'm just a noob undergrad.
GPU memory 16GB required for training.
To train the model on MovieLens100K dataset:
pip install -r requirements.txt python main.py --train_igmc
Once trained (which takes 80 epochs), the model will be saved in the model_dict folder.
