Reformer
An end-to-end deep learning library focused on clear code and speed, used for research and production by Google Brain.
What is Reformer?
Trax is an end-to-end deep learning library developed by Google Brain that focuses on clear code and high performance. It provides a comprehensive workflow from data processing and model building to training and evaluation, supporting a wide range of models including standard architectures like ResNet and Transformer as well as cutting-edge research models. The library is designed to run seamlessly on CPUs, GPUs, and TPUs, offering a straightforward path from experimentation to deployment.
Trax is ideal for deep learning researchers and engineers who need a library that balances readability for rapid experimentation with high performance for production deployment. It is particularly suited for those working on novel model architectures, reinforcement learning, or requiring efficient training on hardware accelerators like TPUs.
Developers choose Trax for its emphasis on readable and simple code without sacrificing performance, enabled by its use of fast math backends like JAX and TensorFlow NumPy. Its unique selling point is providing an end-to-end solution that includes both standard and research models, reinforcement learning algorithms, and flexible data pipelines, all while maintaining ease of use from prototyping to scaling.
Overview
Trax — Deep Learning with Clear Code and Speed
Use Cases
Best For
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.