or `(batch_size, n_classes)` (n_tasks == 1) and the output values are probabilities of each class label.
node_feat, edge_index = data.x, data.edge_index
node_feat = self.embedding(node_feat)
// convolutional layer
for conv in self.conv_layers:
node_feat = conv(node_feat, edge_index)
After Change
This is only returned when self.mode = "classification", the output consists of the
logits for classes before softmax.
node_feats = g.ndata[self.nfeat_name]
out = self.model(g, node_feats)
if self.mode == "classification":
if self.n_tasks == 1:
