jayadratha / pytorchgeometrictutorial

Pytorch Geometric Tutorials

PytorchGeometricTutorial

Hi! We are Antonio Longa and Giovanni Pellegrini, PhD students, and Gabriele Santin, researcher, working between Fondazione Bruno Kessler and the University of Trento, Italy.

This project aims to present through a series of tutorials various techniques in the field of Geometric Deep Learning, focusing on how they work and how to implement them using the Pytorch geometric library, an extension to Pytorch to deal with graphs and structured data, developed by @rusty1s.

You can find our video tutorials on Youtube and at our official website here.

Feel free to join our weekly online tutorial! For more details, have a look at the official website.

Tutorials:

• Tutorial1: What is Geometric Deep Learning? -

• Tutorial2: PyTorch basics. -

• Tutorial3: Graph Attention Network GAT. -

• Tutorial4: Convolutional Layers - Spectral methods. -

• Tutorial5: Aggregation Functions in GNNs. -

• Tutorial6: Graph Autoencoders and Variational Graph Autoencoders. -

• Tutorial7: Adversarially regularized GAE and VGAE. -

• Tutorial8: Graph Generation.

• Tutorial9: Recurrent Graph Neural Networks. -

• Tutorial10: DeepWalk and Node2Vec (Theory).

• Tutorial11: DeepWalk and Node2Vec (Practice). -

• Tutorial12: Edge analysis. -

• Tutorial13: Metapath2vec. -

• Tutorial14: Data handling in Pyg (part 1) -

• Tutorial15: Data handling in Pyg (part 2) -

• Tutorial16: Special guest talk - Matthias Fey

• Tutorial17: Special guest talk - Sergei Ivanov

• Tutorial18: Graph pooling: DIFFPOOL. -

Installation of PyG:

In order to have running notebooks in Colab, we use the following installation commands:

!pip install torch-scatter -f https://data.pyg.org/whl/torch-1.9.0+cu111.html
!pip install torch-sparse -f https://data.pyg.org/whl/torch-1.9.0+cu111.html
!pip install torch-geometric

These version are tested and running in Colab. If instead you run the notebooks on your machine, have a look at the PyG's installation instructions to find suitable versions.