deepmd.pt_expt.loss.tensor

Classes

TensorLoss

Loss on local and global tensors (e.g. dipole, polarizability).

Module Contents

class deepmd.pt_expt.loss.tensor.TensorLoss(tensor_name: str, tensor_size: int, label_name: str, pref_atomic: float = 0.0, pref: float = 0.0, enable_atomic_weight: bool = False, **kwargs: Any)

Bases: deepmd.dpmodel.loss.loss.Loss

Loss on local and global tensors (e.g. dipole, polarizability).

Parameters:

tensor_namestr

The name of the tensor in model predictions.

tensor_sizeint

The size (dimension) of the tensor.

label_namestr

The name of the tensor in labels.

pref_atomicfloat

The prefactor of the weight of atomic (local) loss.

preffloat

The prefactor of the weight of global loss.

enable_atomic_weightbool

If true, atomic weight will be used in the loss calculation.

**kwargs

Other keyword arguments.

tensor_name

tensor_size

label_name

local_weight = 0.0

global_weight = 0.0

enable_atomic_weight = False

has_local_weight

has_global_weight

call(learning_rate: float, natoms: int, model_dict: dict[str, deepmd.dpmodel.array_api.Array], label_dict: dict[str, deepmd.dpmodel.array_api.Array], mae: bool = False) → tuple[deepmd.dpmodel.array_api.Array, dict[str, deepmd.dpmodel.array_api.Array]]

Calculate loss from model results and labeled results.

property label_requirement: list[deepmd.utils.data.DataRequirementItem]

Return data label requirements needed for this loss calculation.

serialize() → dict

Serialize the loss module.

classmethod deserialize(data: dict) → TensorLoss

Deserialize the loss module.