YH-learning

Results 1 comments of YH-learning

Hi @tkipf, Thanks for your shares. When running the code with the big beta parameter.(i.e., _kl_loss + ent_loss_), the performance is not as good as that using _kl_loss / num_nodes...