I’ve been using D-VAE ([1904.11088] D-VAE: A Variational Autoencoder for Directed Acyclic Graphs) lately. It’s a graph neural network that uses asynchronous message-passing, but its github code is too slow. So I’m planning to make a DGL version of D-VAE to speed it up, but I’m curious if DGL can implement asynchronous messaging?
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.