topobench.loss package

Subpackages

• topobench.loss.dataset package
  • Submodules
  • topobench.loss.dataset.DatasetLoss module
    • DatasetLoss
      • DatasetLoss.forward()
      • DatasetLoss.forward_criterion()
  • Module contents
• topobench.loss.model package
  • Submodules
  • topobench.loss.model.DGMLoss module
    • DGMLoss
      • DGMLoss.forward()
  • topobench.loss.model.GraphMLPLoss module
    • GraphMLPLoss
      • GraphMLPLoss.forward()
      • GraphMLPLoss.get_power_adj()
      • GraphMLPLoss.graph_mlp_contrast_loss()
  • Module contents

Submodules

topobench.loss.base module

Abstract class for the loss class.

class topobench.loss.base.AbstractLoss

Bases: ABC

Abstract class for the loss class.

abstract forward(model_out: dict, batch: torch_geometric.data.Data)

Forward pass.

Parameters:

model_outdict

Dictionary containing the model output.

batchtorch_geometric.data.Data

Batch object containing the batched domain data.

topobench.loss.loss module

Loss module for the topobench package.

class topobench.loss.loss.TBLoss(dataset_loss, modules_losses={})

Bases: AbstractLoss

Defines the default model loss for the given task.

Parameters:

dataset_lossdict

Dictionary containing the dataset loss information.

modules_lossesAbstractLoss, optional

Custom modules’ losses to be used.

forward(model_out: dict, batch: torch_geometric.data.Data)

Forward pass of the loss function.

Parameters:

model_outdict

Dictionary containing the model output.

batchtorch_geometric.data.Data

Batch object containing the batched domain data.

Returns:

dict

Dictionary containing the model output with the loss.

Module contents

This module implements the loss functions for the topobench package.

class topobench.loss.TBLoss(dataset_loss, modules_losses={})

Bases: AbstractLoss

Defines the default model loss for the given task.

Parameters:

dataset_lossdict

Dictionary containing the dataset loss information.

modules_lossesAbstractLoss, optional

Custom modules’ losses to be used.

forward(model_out: dict, batch: torch_geometric.data.Data)

Forward pass of the loss function.

Parameters:

model_outdict

Dictionary containing the model output.

batchtorch_geometric.data.Data

Batch object containing the batched domain data.

Returns:

dict

Dictionary containing the model output with the loss.