Thang Duong

@unsuthee have you implemented it? I have changed the loss function to Eq. 1 but the embeddings do not make sense: connected nodes do not have close embeddings. They are...

My understanding is that the class_map is needed but it is not used for unsupervised training. You can assign random classes to the nodes.

Try adding this before every writer.add_image() ` data = data.transpose(2,0,1)` Or else, you can specify the image format based on this https://github.com/lanpa/tensorboardX/blob/master/tensorboardX/writer.py#L355