Boost Your Crystal Model with Denoising Pre-training

[OpenReview] [Project Page]

News

Stay updated with the latest milestones of our work:

• 🌟 Featured at ICML AI4Science Workshop 2024
  "Boost Your Crystal Model with Denoising Pre-training"

• 🌟 Presented at AAAI Conference 2025
  "A Denoising Pre-training Framework for Accelerating Material Discovery"

Dataset

The dataset used for pre-training can be found at GNoME.

Training and Prediction

You can train and test the model with the following commands:

conda env create -f DPF.yaml
conda activate DPF
cd matformer
bash pretrain.sh

For training your own custom models, you only need to replace the model with your own.

Citation

Please cite our paper if you find the code helpful.

@inproceedings{DPF_ICML,
  title={Boost Your Crystal Model with Denoising Pre-training},
  author={Shuaike Shen and Ke Liu and Muzhi Zhu and Hao Chen},
  booktitle={ICML 2024 AI for Science Workshop},
  year={2024},
  url={https://openreview.net/forum?id=u2qYzRRg02}
}

@inproceedings{DPF_AAAI,
  title={A Denoising Pre-training Framework for Accelerating Novel Material Discovery},
  author={Shen, Shuaike and Liu, Ke and Zhu, Muzhi and Chen, Hao},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={39},
  number={27},
  pages={28368--28376},
  year={2025}
}

@inproceedings{ijcai2022p708,
  title     = {S2SNet: A Pretrained Neural Network for Superconductivity Discovery},
  author    = {Liu, Ke and Yang, Kaifan and Zhang, Jiahong and Xu, Renjun},
  booktitle = {Proceedings of the Thirty-First International Joint Conference on
               Artificial Intelligence, {IJCAI-22}},
  publisher = {International Joint Conferences on Artificial Intelligence Organization},
  editor    = {Lud De Raedt},
  pages     = {5101--5107},
  year      = {2022},
  month     = {7},
  doi       = {10.24963/ijcai.2022/708},
  url       = {https://doi.org/10.24963/ijcai.2022/708},
}

Acknowledgement

This repo is built upon the previous work ALIGNN and MatFormer. The original idea comes from S2SNet.

Contact

If you have any questions, please contact me at [email protected]