gammagl.models.HardGATModel

class HardGATModel(feature_dim, hidden_dim, num_class, heads, drop_rate, k, num_layers, name=None)

The graph hard attentional operator from the “Graph Representation Learning via Hard and Channel-Wise Attention Networks” paper.

Parameters:

• feature_dim (int) – input feature dimension.

• hidden_dim (int) – hidden dimension.

• num_class (int) – number of classes.

• heads (int) – number of attention heads.

• drop_rate (float) – dropout rate.

• k (int) – number of neighbors to attention.

• name (str, optional) – model name.

forward(x, edge_index, num_nodes)

Define the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.