LINKX PaddlePaddle Implementation

This repository contains the PaddlePaddle implementation of LINKX (Large Scale Learning on Non-Homophilous Graphs, NeurIPS 2021).

Environment Requirements

To run the code, you need to install the following:

• paddle - PaddlePaddle deep learning framework
• pgl - Paddle Graph Learning, a graph deep learning framework developed by Baidu

Datasets

The model has been tested on the following dataset:

• Wiki - For more details regarding the dataset, visit the Non-Homophily Large-Scale GitHub repository.

Instructions

To train the model, run the main.py script. You can modify the following parameters as needed:

python main.py --train_batch cluster \
               --no_mini_batch_test \
               --batch_size 10000 \
               --num_parts 100 \
               --cluster_batch_size 1 \
               --saint_num_steps 5 \
               --test_num_parts 10

How to Run

First, ensure you have installed the required environments as mentioned above. You can then clone this repository and navigate to the directory where the main.py is located. After that, you can run the training process with the desired parameters.