Skip to content

v0.1.0

Choose a tag to compare

@chrbrunk chrbrunk released this 02 Jun 10:28
· 12 commits to main since this release

Initial release with the following features:

  • Implemented model architectures: MACE, NequIP and ViSNet
  • Dataset preprocessing
  • Training of MLIP models
  • Batched inference with trained MLIP models
  • MD simulations with MLIP models using JAX-MD and ASE simulation backends
  • Energy minimizations with MLIP models using the same simulation backends
  • Fine-tuning of pre-trained MLIP models (only for MACE)