Update

    --------

    >>> import deepchem as dc

    >>> import pymatgen as mg

    >>> import dgl

    >>> from deepchem.models import GCN

    >>> lattice = mg.Lattice.cubic(4.2)

    >>> structure = mg.Structure(lattice, ["Cs", "Cl"], [[0, 0, 0], [0.5, 0.5, 0.5]])

    >>> featurizer = dc.feat.CGCNNFeaturizer()

    >>> cgcnn_graph = featurizer.featurize([structure])[0]

    >>> cgcnn_graph.num_node_features

    92

    >>> cgcnn_dgl_graph = cgcnn_graph.to_dgl_graph()

    >>> model = GCN(in_node_dim=92, hidden_node_dim=92, num_gnn_layers=2)

    >>> # Call model.eval as batch norm is implemented

    >>> model.eval()

    >>> model(cgcnn_dgl_graph)

    >>> smiles = ["C1CCC1", "C1=CC=CN=C1"]

    >>> featurizer = dc.feat.MolGraphConvFeaturizer()

    >>> graphs = featurizer.featurize(smiles)

    >>> print(type(graphs[0]))

    <class 'deepchem.feat.graph_data.GraphData'>

    >>> dgl_graphs = [graphs[i].to_dgl_graph() for i in range(len(graphs))]

    >>> # Batch two graphs into a graph of two connected components

    >>> batch_dgl_graph = dgl.batch(dgl_graphs)

    >>> model = GCN(n_tasks=1, number_atom_features=30, mode='regression')

    >>> model(batch_dgl_graph)

    References

    ----------

            This is only returned when self.mode = 'classification', the output consists of the

            logits for classes before softmax.

        """

        node_feats = g.ndata.pop(self.nfeat_name)

        node_feats = g.ndata[self.nfeat_name]

        out = self.model(g, node_feats)

        if self.mode == 'classification':