Psiflow is a modular and scalable library for developing interatomic potentials. It interfaces popular trainable interaction potentials with quantum chemistry software and is designed to support computational workflows on hundreds or thousands of nodes. Psiflow is designed as an end-to-end framework; it can orchestrate all computational components between an initial atomic structure and the final trained potential. In particular, it implements a variety of active learning algorithms which allow for efficient exploration of the system's phase space without requiring ab initio molecular dynamics.
Its features include:
- active learning algorithms with enhanced sampling using PLUMED
- Weights & Biases logging for easy monitoring and analysis
- periodic (CP2K) and nonperiodic (PySCF, NWChem) systems
- efficient GPU-accelerated molecular dynamics using OpenMM
- the latest equivariant potentials such as MACE and NequIP
Execution is massively parallel and powered by Parsl, a parallel execution library which supports a variety of execution resources including clouds (e.g. Amazon Web Services, Google Cloud), clusters (e.g. SLURM, Torque/PBS, HTCondor) and even container orchestration systems (e.g. Kubernetes). While psiflow exposes an intuitive and concise API for defining complex molecular simulation workflows in a single Python script, Parsl ensures that the execution is automatically offloaded to arbitrarily large amounts of compute resources. Visit the documentation for more details.
PREVIEW: click here for a sneak peak into a few psiflow-generated training sets for the formic acid proton transfer!
Check out this seven-minute introduction which was recorded at ParslFest 2023:
