Open mmbaghi opened 7 months ago
I faced the same problem. Have you solved it?
I faced the same problem. Have you solved it?
Hi, the torch code simplifies the model (e.g. use simple convolutional HGNN, use simple alignment module). You can refer to the propagate function in the original tensorflow code for the implementation of the hypergraph transformer.
Hello I read your article. In the article, you said that you used Hypergraph Transformer, but I did not find the Transformer section in my code.
You only wrote the last function in TensorFlow 2 function ( lightSelfAttention, selfAttention ), but it was not used anywhere. Please guide what should I do to implement Transformer?
And have you used gcn instead of lightgcn?