This is a work-in-progress rework of the TransFM code here: https://github.com/rpasricha/TransFM

For a high level demonstration of the code see here: https://colab.research.google.com/drive/1oAq1cGgo7XttF0Zpgr55zF8gtT4VdNEE?usp=sharing

More feature demos and documentation will be added soon.

While the original code was from the above repository, the changes here are extensive and include:

-Changing Tensorflow version to 2.x

-Adding functionality for outputting predections

-Memory Optimizations most notably the ability to train in segments

-Adding option to use regional data, limiting users to interacting with only certain items

-Ability to save and load model in a way that preserves user data

-Ability to add new items and users to a partially trained model

-Option to only use train/test, allowing more users to be covered by model

-Model seeding allows for determininsitic results

-Args object allows for easier debugging and saving of run parameters

-Added pyton notebook that can be used in google colab to quickly demo model