📖 Full documentation: https://atomgptlab.github.io/alignn/
- Introduction
- Installation
- Examples
- Pre-trained models
- JARVIS-ALIGNN webapp
- ALIGNN-FF & ASE Calculator
- Peformances on a few datasets
- Useful notes
- References
- How to contribute
- Correspondence
- Funding support
The Atomistic Line Graph Neural Network (paper) introduces a graph convolution layer that explicitly models both two- and three-body interactions in atomistic systems. The ALIGNN-FF variant (paper) extends this to a force-field for structurally and chemically diverse systems across 89 elements.
See docs/index.md for the full introduction.
See docs/installation.md for conda, GitHub, and pip installation methods.
See docs/training/ for dataset format and training examples:
- Dataset format
- Single-output regression
- Classification
- Multi-output regression
- Force-field training
- Multi-GPU training
Google Colab notebooks are linked from docs/index.md.
See docs/pretrained/:
See docs/usage/webapps.md. Direct links: AtomGPT ALIGNN app, ALIGNN-FF app.
See docs/usage/ase-calculator.md for example usage.
See docs/performance.md for benchmark tables on JARVIS-DFT, Materials Project, QM9, hMOF, qMOF, OpenCatalyst, and other datasets. Also see JARVIS-Leaderboard.
See docs/notes.md for common pitfalls and FAQs.
See docs/references.md for the publication list.
See Contribution instructions and docs/contributing.md.
Please report bugs as GitHub issues or email drkamal@jhu.edu.
Please see Code of conduct.