A PyTorch implementation of "Graph Wavelet Neural Network" (ICLR 2019)
Geek Repo:Geek Repo
Github PK Tool:Github PK Tool
yyysyyy opened this issue a year ago · comments
I found that in the train_test_split function in gwnn.py, you divide 80% of the nodes in the cora data set into training sets for training. Is this in line with common sense? Or I misunderstood or your code is wrong?