PotentialLearning.jl
PotentialLearning.jl copied to clipboard
cesmix-mit
An open source Julia library for active learning of interatomic potentials in atomistic simulations of materials. It incorporates elements of bayesian inference, machine learning, differentiable progr...
Output of status on Julia 1.9 ``` (PotentialLearning) pkg> st Project PotentialLearning v0.2.2 Status `~/src/PotentialLearning/Project.toml` [a963bdd2] AtomsBase v0.3.5 [aaaa29a8] Clustering v0.15.7 [2673d5e8] Determinantal v0.2.0 [31c24e10] Distributions v0.25.107 [587475ba] Flux v0.14.15...
Description [here](https://docs.google.com/document/d/1SWAanEWQkpsbr2lqetMO3uvdX_QK-Z7dwrgPaM1Dl0o/edit).
Package refactor integrating the code of the following sources: - Latest developments in atomistic composable worflows: https://github.com/cesmix-mit/AtomisticComposableWorkflows/tree/master/ACE - Active learning branch: https://github.com/cesmix-mit/PotentialLearning.jl/tree/active_learning_refactor - Old refactor branch: https://github.com/cesmix-mit/PotentialLearning.jl/tree/refactor
Make this possible: `] add PotentialLearning.jl`
See https://cesmix-mit.github.io/LAMMPS.jl/dev/generated/fitting_snap/
Three variants of CUR are implemented in PotentialLearning.jl: LinearTimeCUR, DEIMCUR, and LSCUR.
Metadata
Owner
Metadata
An open source Julia library for active learning of interatomic potentials in atomistic simulations of materials. It incorporates elements of bayesian inference, machine learning, differentiable progr...