-
Notifications
You must be signed in to change notification settings - Fork 0
Home
We aim to establish a comprehensive and interoperable ecosystem tailored for differentiable physical modeling. The initial focus is on streamlining neural network potentials (NNPs) for efficient and accurate free energy calculations. To achieve this, we develop three packages intended to be used together but also useful as stand-alone: the auditorium(https://github.com/choderalab/auditorium) package will provide our central testing and benchmarking ground for NNPs, as well as the modelforge package, which contains the implementation, training, distribution, storage, and application of NNPs designed for molecular simulations.
The chiron package will provide the Markov Chain Monte Carlo (MCMC) state sampler and Molecular Dynamics engine, chiron.