SAGPool

SAGPool copied to clipboard

burger-han

in train epoch: why optimizer.zero_grad() after optimizer.step()? Does it matter? It's usually optimizer.step()--loss.backward()--optimizer.step()

littleWangyu

Batch operation

1

Hi, thank for sharing the codes. In the networks.py, what does the variable "data.batch" denote in the forward method? I have looked up to pyg document, and found neither Data...

lonelygoatherd

Reproducibility of Table 3

9

Hi, I am running your code several time on the DD dataset and obtain different results than the one you present on Table 3 of your paper. In particular I...

SilvioGiancola

Hi，your work is so interesting! In the experiments, you said "we evaluated the pooling methods over 20 random seeds using 10-fold cross validation. " Could you tell me what random...

xyhappy

'str' object has no attribute 'fotmat'

3

https://github.com/inyeoplee77/SAGPool/blob/4988acddc9319681826fb03fe6141701e9714693/main.py#L100 The 100-row code has a spelling problem. There should be 'format' instead of 'fotmat'.

lfywork

Hi! The FRANKENSTEIN Dataset does not have any node attributes by default. Was SAGPool trained on the additional node attributes using `node_attr=True` in the `TUDataset` constructor or on some other...

ekagra-ranjan

torch_geometric?

1

Hello, thanks for the amazing work, I see you import torch_geometric, so is this code based on rusty1s' pytorch_geometric?

Yfhu1103

run code on PROTEINS dataset error

4

Hi, thanks for your code. But when I run python main.py --dataset PROTEINS, there is an error as follows: ```Validation loss:0.513576095168655 accuracy:0.7567567567567568 Traceback (most recent call last): File "main.py", line...

ColdenChan

@inyeoplee77 Hi, I am wondering if the stats provided in the paper under proteins column is for Proteins, proteins_full, or concatenation, node-label features, etc.?

choltz95