Title

Docs for fande models.

start out with fande.model.....

Standard Models

Model 1

class fande.models.ModelForces(*args: Any, **kwargs: Any)

Batch independent multioutput exact gp model.

\[\begin{equation} k_{i}(\mathbf{x}, \mathbf{x'}) \end{equation}\]

Note

Simple note.

configure_optimizers()

defines model optimizer

forward(input_)

Compute prediction.

predict_forces(input_)

Just prediction step with additional reshape taking into account the atomic group

Returns:

reshaped prediction

predict_step(batch, batch_idx: int, dataloader_idx: int = None)

wrapper around lightning function

test_step(batch, batch_idx)

Compute testing loss.

training_step(batch, batch_idx: int, dataloader_idx: int = None)

Compute training loss.

class fande.models.ModelEnergies(*args: Any, **kwargs: Any)

configure_optimizers()

defines model optimizer

forward(input_)

Compute prediction.

test_step(batch, batch_idx)

Compute testing loss.

training_step(batch, batch_idx: int, dataloader_idx: int = None)

Compute training loss.