danielegrattarola/spektral
A Python library for building graph neural networks (GNNs) using Keras and TensorFlow 2.
Not currently ranked — collecting fresh signals.
star history
Spektral provides a Python framework for creating graph neural networks, implementing popular GNN layers including GCN, GraphSAGE, GAT, GIN, and various pooling methods. Built on Keras and TensorFlow 2, it offers utilities for graph representation, manipulation, and transformation to support graph-based deep learning tasks such as node classification, molecular property prediction, and link prediction.
Frequently asked
- What is danielegrattarola/spektral?
- A Python library for building graph neural networks (GNNs) using Keras and TensorFlow 2.
- Is spektral open source?
- Yes — danielegrattarola/spektral is open source, released under the MIT license.
- What language is spektral written in?
- danielegrattarola/spektral is primarily written in Python.
- How popular is spektral?
- danielegrattarola/spektral has 2.4k stars on GitHub.
- Where can I find spektral?
- danielegrattarola/spektral is on GitHub at https://github.com/danielegrattarola/spektral.