mir-group/allegro
A deep learning model for predicting interatomic potentials in molecular dynamics and computational chemistry simulations.
Not currently ranked — collecting fresh signals.
star history
Allegro implements an E(3)-equivariant machine learning interatomic potential model as an extension package for the NequIP framework. It uses PyTorch to train neural network models that predict atomic forces and energies for molecular dynamics simulations. The model architecture incorporates geometric symmetries specific to atomic systems and integrates with LAMMPS for running simulations at scale.
Frequently asked
- What is mir-group/allegro?
- A deep learning model for predicting interatomic potentials in molecular dynamics and computational chemistry simulations.
- Is allegro open source?
- Yes — mir-group/allegro is open source, released under the MIT license.
- What language is allegro written in?
- mir-group/allegro is primarily written in Python.
- How popular is allegro?
- mir-group/allegro has 490 stars on GitHub.
- Where can I find allegro?
- mir-group/allegro is on GitHub at https://github.com/mir-group/allegro.