chao1224/GraphMVP

I cannot use other gnn models in your code

nickspark opened this issue · 2 comments

when I set
gnn_type = 'graphsage'
to use graphsage instead of gin, I got the following errors, do you have any idea about it?
MessagePassing.propagate only supports torch.LongTensor of shape [2, num_messages] or torch_sparse.SparseTensor for argument edge_index.

Hi @nickspark , our current implementation is using GIN only.
For the GraphSAGE, we followed this repo.
Yet, I will check this and get back to you soon.

Hi @nickspark ,
I updated the GNN with the latest pyg version. Feel free to check out.