I have trained a GCN to get the node embedings for karate dataset.
The codes, revised model files and the result pic are listed.

I have trained a GCN to get the node embedings for karate dataset.
The codes, revised model files and the result pic are listed.

Hi,thanks for your sharing.
I run your code to create two file: karate.cites and karate.content, it success. However when I runed train_karate.py, it made this mistake:
RuntimeError: Expected object of type torch.cuda.LongTensor but found type torch.LongTensor for argument #3 'index'

This question point to this sentence:
loss = criterion(torch.index_select(output, 0, indices),torch.LongTensor([0,1,2,3]))
(I try to change it to: loss = criterion(torch.index_select(output, 0, indices),torch.cuda.LongTensor([0,1,2,3])) , but it doesn't work)

The complete error is as follow:
/home/wtu/anaconda3/envs/py/lib/python3.5/site-packages/torch/nn/functional.py:995: UserWarning: nn.functional.tanh is deprecated. Use torch.tanh instead.
warnings.warn("nn.functional.tanh is deprecated. Use torch.tanh instead.")
/home/wtu/project/pygcn_data/models2.py:22: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.
return F.softmax(x)
Traceback (most recent call last):
File "/home/wtu/project/pygcn_data/train_karate.py", line 92, in
train(epoch)
File "/home/wtu/project/pygcn_data/train_karate.py", line 80, in train
loss = criterion(torch.index_select(output, 0, indices),torch.LongTensor([0,1,2,3]))
RuntimeError: Expected object of type torch.cuda.LongTensor but found type torch.LongTensor for argument #3 'index'

looking forward to your reply

Try running it on CPU only or out that tensor explicitly on GPU.

Thanks for your reply, I work it out.
I change "loss = criterion(torch.index_select(output, 0, indices),torch.LongTensor([0,1,2,3]))" to
"loss = criterion(torch.index_select(output, 0, indices),torch.cuda.LongTensor([0,1,2,3]))"
and change "indices = torch.LongTensor([0, 16,18,24])" to
"indices = torch.cuda.LongTensor([0, 16,18,24])"
Then success~