torch_geometric.nn.conv.FastFiLMConv

class FastFiLMConv(in_channels: Union[int, Tuple[int, int]], out_channels: int, num_relations: int = 1, nn: Optional[Callable] = None, act: Optional[Callable] = ReLU(), aggr: str = 'mean', is_sorted: bool = False, **kwargs)

Bases: MessagePassing

See FiLMConv. Main difference is parrallelizing linear layers at the cost of more memory usage. For optimal performance, edge_index should be sorted by edge_type.

forward(x: Union[Tensor, Tuple[Tensor, Tensor]], edge_index: Union[Tensor, SparseTensor], edge_type: Optional[Tensor] = None) → Tensor

Runs the forward pass of the module.

reset_parameters()

Resets all learnable parameters of the module.