graphnet.models.gnn.dynedge_jinst module

Implementation of the exact DynEdge architecture used in [2209.03042].

Author: Rasmus Oersoe

class graphnet.models.gnn.dynedge_jinst.DynEdgeJINST(*args, **kwargs)[source]

Bases: GNN

DynEdge (dynamical edge convolutional) model used in [2209.03042].

Construct DynEdgeJINST.

Parameters:

  • nb_inputs (int) – Number of input features.

  • nb_outputs – Number of output features.

  • layer_size_scale (int, default: 4) – Integer that scales the size of hidden layers.

  • args (Any)

  • kwargs (Any)

Return type:

object

forward(data)[source]

Apply learnable forward pass.

Return type:

Tensor

Parameters:

data (Data)