gmx::INNPotModel Class Reference

#include <gromacs/applied_forces/nnpot/nnpotmodel.h>

Inheritance diagram for gmx::INNPotModel:

Description

NNPot Module.

Abstract class for a neural network potential model. Inherit from this class to implement a specific neural network backend/framework. Refer to torchmodel.h for an example implementation.

Public Member Functions
virtual void	initModel ()=0
	Initialize the neural network model.
virtual void	evaluateModel ()=0
	call inference on NN model
virtual void	getOutputs (std::vector< int > &, gmx_enerdata_t &, const ArrayRef< RVec > &, bool)=0
	retrieve NN model outputs
virtual void	setCommRec (const t_commrec *)=0
	set communication record for possible communication of input/output data between ranks
virtual void	prepareAtomPositions (std::vector< RVec > &)=0
	Prepare inputs for NN model. More...
virtual void	prepareAtomNumbers (std::vector< int > &)=0
virtual void	prepareBox (matrix &)=0
virtual void	preparePbcType (PbcType &)=0

Member Function Documentation

virtual void gmx::INNPotModel::prepareAtomPositions ( std::vector< RVec > &  )

pure virtual

Prepare inputs for NN model.

Currently supported inputs:

• atom positions (std::vector<RVec>): atomic positions
• atom numbers (std::vector<int>): atomic numbers
• box (matrix): simulation box vectors
• pbc type (PbcType): boolean flags for periodic boundary conditions in x, y, z

Implemented in gmx::TorchModel.

---

The documentation for this class was generated from the following file:

• src/gromacs/applied_forces/nnpot/nnpotmodel.h