Awesome

RecBole-TRM

RecBole-TRM is a library built upon PyTorch and RecBole for reproducing and developing recommendation algorithms based on Transformers (TRMs). Our library includes algorithms covering two major categories:

• Sequential Recommendation: TiSASRec, SSE-PT, LightSANs, gMLP, CORE

Highlights

• Easy-to-use API: Our library shares unified API and input (atomic files) as RecBole.
• Fair reproducibility and comparison: Our library provides fair reproducibility and comparison in a systematic mechanism.
• Extensive Transformer library: Our library provides extensive API based on common Transformer layers, one can further develop new models easily based on our library.

Requirements

recbole>=1.0.0
pyg>=2.0.4
pytorch>=1.7.0
python>=3.7.0

Quick-Start

With the source code, you can use the provided script for initial usage of our library:

python run_recbole_trm.py

If you want to change the models or datasets, just run the script by setting additional command parameters:

python run_recbole_trm.py -m [model] -d [dataset]

Implemented Models

We list currently supported models according to category:

Sequential Recommendation:

• TiSASRec from Li et al.: Time Interval Aware Self-Attention for Sequential Recommendation (WSDM 2020).
• SSE-PT from Wu et al.: SSE-PT: Sequential Recommendation Via Personalized Transformer (RecSys 2020).
• LightSANs from Fan et al.: Lighter and Better: Low-Rank Decomposed Self-Attention Networks for Next-Item Recommendation (SIGIR 2021).
• gMLP from Liu et al.: Pay Attention to MLPs (NeurIPS 2021).
• CORE from Hou et al.: CORE: Simple and Effective Session-based Recommendation within Consistent Representation Space (SIGIR 2022).

Experiments

For more details about experiments including the hyper-parameters of the implemented models, you can refer to [link].

The Team

RecBole-TRM is developed and maintained by members from RUCAIBox, the main developers are Wenqi Sun (@wenqisun) and Xinyan Fan (@BELIEVEfxy).

Acknowledgement

The implementation is based on the open-source recommendation library RecBole.

Please cite the following paper as the reference if you use our code or processed datasets.

@inproceedings{zhao2021recbole,
  title={Recbole: Towards a unified, comprehensive and efficient framework for recommendation algorithms},
  author={Wayne Xin Zhao and Shanlei Mu and Yupeng Hou and Zihan Lin and Kaiyuan Li and Yushuo Chen and Yujie Lu and Hui Wang and Changxin Tian and Xingyu Pan and Yingqian Min and Zhichao Feng and Xinyan Fan and Xu Chen and Pengfei Wang and Wendi Ji and Yaliang Li and Xiaoling Wang and Ji-Rong Wen},
  booktitle={{CIKM}},
  year={2021}
}