search

THUDM / GraphMAE

GraphMAE: Self-Supervised Masked Graph Autoencoders in KDD'22
478 stars 75 forks source link

Code about Reddit Dataset #43

Closed yolandalalala closed 1 year ago

yolandalalala commented 1 year ago

Hi, thanks for releasing the code of your excellent work in Graph Neural Networks!

I got a question regarding the Reddit Dataset. Your results on Reddit presented in your paper shows promising result and fabulous generalizability of GraphMAE, so that I hope to reproduce that result myself. However, if I understand correctly, data loader for Reddit is not included. I assume there is some special accommodation on Reddit is needed to train the GraphMAE because Reddit has 23k nodes and 11m edges. Could you please kindly provide your code for processing Reddit Dataset? Thank you so much for your consideration.

THINK2TRY commented 1 year ago

@yolandalalala Thanks for your attention to GraphMAE! In fact, we conduct full-batch training for Reddit dataset as well as Cora without any mini-batch data loader. A direct way is to use a GPU with larger memory so you can run the program and reproduce the results. If you would to use a dataloader for mini-batch training, you can write a customized one using the dataloaders provided in DGL or refer to our implementation here.