Graph Attention Neural Networks for Mapping Materials and Molecules
Our training is carried out on a system with 10 CPU cores, 32 GB of RAM, and an NVIDIA GeForce GTX 2080 Ti GPU. Check requirements.txt for other dependencies.
We provide an implementation for the QMOF experiment.
- Data
(a) Download database file from https://github.com/arosen93/QMOF or https://figshare.com/articles/dataset/QMOF_Database/13147324 (b) Find 'opt-geometries.xyz' from the downloaded file and place it in /GANN/data/ (c) cd /GANN/data/ (c) Run the python script to process into structure files with "python xyz_to_cifs.py"
- Training
To train the model in the paper, run these commands: cd /GANN python main.py --batch-size 32 --train-ratio 0.8 --val-ratio 0.1 --test-ratio 0.1 --workers 1 --epochs 200 --print-freq 1 data > log.out
- Evaluation
All results are available in the file 'output'.