Learning Superconductivity from Ordered and Disordered Material Structures (NeuIPS 2024)

A transformer-based GNN model for learning superconductivity from ordered and disordered crystal structures. Implementation codes for Learning Superconductivity from Ordered and Disordered Material Structure.
[Paper]

Requirement

The important packages are presented as follows:

e3nn                  0.4.4
numpy                 1.22.4
pymatgen              2023.2.28
scipy                 1.8.1
timm                  0.4.12
torch                 1.10.2+cu111
torch-cluster         1.6.0
torch-geometric       2.2.0
torch-scatter         2.0.9
torch-sparse          0.6.13
torch-spline-conv     1.2.1
torchaudio            0.10.2 
torchmetrics          0.8.2
torchvision           0.11.3+cu111
tqdm                  4.65.0 

Dataset

The dataset is undered datasets/SuperCon/cif/ and the Tc values are saved in datasets/SuperCon/df_all_data1202.csv. More details can be found in datasets/SuperCon/README.

Example

Some tests on data processing, modeling and inference are given in the examples/test.py You can run the test with the following command and determine if your environment is installed correctly:

    python test.py

Training

All the training scripts are under scripts/SuperCon/ . The input data will be divided into 10-fold before training, so you can train according to the number of folds you want to run. For example:

    sh scripts/SuperCon/train_[FOLD].sh

If you want to run all the folds at once, you can use the following command:

    sh scripts/SuperCon/train_all.sh

Inference

After training, all models will be saved in best_models/. You can use these *_save.pt files for inference with the following commands:

    sh scripts/infer/infer.sh

The results of inference will be saved in pred.json.

Citation

Please consider citing our work if you find it helpful:

@inproceedings{chenlearning,
  title={Learning Superconductivity from Ordered and Disordered Material Structures},
  author={Chen, Pin and Peng, Luoxuan and Jiao, Rui and Mo, Qing and Zhen, WANG and Huang, Wenbing and Liu, Yang and Lu, Yutong},
  booktitle={The Thirty-eight Conference on Neural Information Processing Systems Datasets and Benchmarks Track},
  year={2024}
}