This project is an implementation of GraphLoG, a method for self-supervised graph representation learning with local and global structure. The repo provides the pre-training and …
Keywords: graph representation learning, unsupervised learning, self-supervision, graph classification, geometric deep learning, machine learning, EM, Python, PyTorch, PyG
Reference implementation of memory-based graph neural networks. This work introduces an efficient memory layer to jointly learn representations and coarsen the input graphs.
Keywords: graph memory net, MGN, Memory-Based Graph Networks, graph machine learning, geometric deep learning, , MemGNN, graph classification, graph neural networks
Graph Neural Architecture Search (GraphNAS for short) enables automatic design of the best graph neural architecture based on reinforcement learning. Specifically, GraphNAS first uses a …
Keywords: neural architecture search, graph neural networks, AutoML, NAS, machine learning, Python, PyTorch
Graphormer is a general-purpose deep learning backbone for molecular modeling. Graphormer is a deep learning package that allows researchers and developers to train custom models …
Keywords: graph representation learning, graph machine learning, molecule science, transformer, machine learning, Python, PyTorch
The reference implementation for the GraphSAGE algorithm using PyTorch. For small graphs, It is more efficient that the corresponding TensorFlow implementation.
Keywords: graph neural network, graph sampling and aggregation, scalable GNN, PyTorch, machine learning, Python
The reference implementation of the GraphSAGE algorithm in Tensorflow.
Keywords: graph neural network, graph sampling and aggregation, scalable GNN, TensorFlow, machine learning, Python
GraphStorm is a graph machine learning (GML) framework designed for enterprise use cases. It simplifies the development, training and deployment of GML models on industry-scale …
Keywords: graph machine learning, graph neural networks, representation learning, scalability, GPU, NVIDIA, DGL, Python, PyTorch
The reference implementation of the Graph Transformer architecture that extends the Transformer architecture used in Natural Language Processing (NLP) to graph-structured data. It extends Transformer …
Keywords: graph representation learning, transformer architecture, attention, graph neural networks, GNN, Python, PyTorch
Graph Transformer Networks (GTNs)
The reference implementation of Graph Transformer Networks for node classification in heterogeneous graphs.
Keywords: graph representation learning, transformer, attention, heterogenous graph, node classification, GTN, Python, PyTorch
Graphtyper is a graph-based variant caller capable of genotyping population-scale short read data sets. It represents a reference genome and known variants of a genomic …
Keywords: genomics, bioinformatics, variant calling, graph, pangenome