How do I run CondGEN experiments in your paper?

[xiangsheng1325]

I have trained a CondGEN model with the default settings in options.py. But I can't make the trained model perform as well as yours scriped in the paper. And I also find some differences between your code and paper, such as the number of GCN layers in the encoder, which is 1 in your code but 2 in your paper. So how can I have the hyperparameters setting to reproduce your experiment results? Thanks!