diffpool
diffpool copied to clipboard
RexYing
Thank you very much for your interesting paper and code. I have a concern regarding the naming. Instead of pooling, I think what you proposed is a fully connected layer...
I am a little bit confused by the batch normalization implementation here: everytime I run bn a `self.bn(x)` is used, where a new bn layer is created, i.e., `bn_module =...
hello, I edited parser.set_defaults{'method'='soft-assign' }, then error happened on the SoftPoolingGcnEncoder. So I have some questions here: Whether nn.Linear() should add .cuda() on 90th line? Whether self.assign_pred should change into...
Hi, How do you deal with input graph with different size, since once the assignment matrix S is learned,its size is fixed. How to solve different graphs with different number...
Hi, According to the paper, it said you used the GraphSAGE as the GNN module and stack two layers before diffpool layer. However, in this code implementation, are you using...
I was trying to run the file `example.sh`. The first command runs fine, but it throws an exception on the second one: `$ python -m train --bmname=ENZYMES --assign-ratio=0.1 --hidden-dim=30 --output-dim=30...
why the maximum number of nodes are constrained? and what about other dataset except for D&D, ENZYMES? in `example.sh`, there are only information (of max_node) about two dataset.
Is this the official implementation of the paper | Hierarchical Graph Representation Learning with Differentiable Pooling
I was trying to run the file example.sh. The first command runs fine, but it throws an exception on the second one: $ python -m train --bmname=ENZYMES --assign-ratio=0.1 --hidden-dim=30 --output-dim=30...
