convnet

Implementation of the ConvNet GNN model architecture.

Author: Martin Ha Minh

class graphnet.models.gnn.convnet.ConvNet(*args, **kwargs)[source]

Bases: GNN

ConvNet (convolutional network) model.

Construct ConvNet.

Parameters:

  • nb_inputs (int) – Number of input features, i.e. dimension of input layer.

  • nb_outputs (int) – Number of prediction labels, i.e. dimension of output layer.

  • nb_intermediate (int, default: 128) – Number of nodes in intermediate layer(s).

  • dropout_ratio (float, default: 0.3) – Fraction of nodes to drop.

  • args (Any)

  • kwargs (Any)

Return type:

object

forward(data)[source]

Apply learnable forward pass.

Return type:

Tensor

Parameters:

data (Data)