Sonnet - a library for constructing neural networks by Google's DeepMind
A TensorFlow 2 library providing simple, composable abstractions for machine learning research via the snt.Module concept.
What is Sonnet - a library for constructing neural networks by Google's DeepMind?
Sonnet is a TensorFlow-based neural network library developed by DeepMind, designed to provide simple, composable abstractions for machine learning research. It introduces the `snt.Module` concept as a core building block for constructing self-contained neural network components, enabling flexible and decoupled design for various ML tasks like supervised learning, reinforcement learning, and more.
Machine learning researchers and developers who need a flexible, unopinionated library for building and experimenting with neural networks in TensorFlow, particularly those who want fine-grained control over their model architecture and training process.
Developers choose Sonnet for its simplicity, clarity, and powerful yet unopinionated programming model centered around modules, which allows for full control over training and model design without being tied to a specific framework, making it ideal for research and custom implementations.
Overview
TensorFlow-based neural network library
Use Cases
Best For
- Building custom neural network architectures from scratch
Related Projects
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
