mila-iqia/torch_jax_interop

Simple tools to mix and match PyTorch and Jax - Get the best of both worlds!

35
/ 100
Emerging

This tool helps machine learning engineers and researchers combine elements from PyTorch and JAX within a single project. It allows you to use PyTorch models or functions inside a JAX workflow, or vice-versa, by converting data structures (tensors) and wrapping functions between the two frameworks. The result is that you can leverage the strengths of both, like JAX's performance optimizations and PyTorch's mature ecosystem, without rewriting existing code.

Use this if you are a machine learning practitioner who wants to integrate components built in PyTorch with JAX, or JAX components with PyTorch, to optimize performance or leverage specific framework features.

Not ideal if you are starting a new project from scratch and prefer to commit to a single deep learning framework for your entire codebase.

deep-learning machine-learning-engineering model-development ai-research ml-infrastructure
No Package No Dependents
Maintenance 6 / 25
Adoption 7 / 25
Maturity 16 / 25
Community 6 / 25

How are scores calculated?

Stars

37

Forks

2

Language

Python

License

MIT

Last pushed

Jan 08, 2026

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/mila-iqia/torch_jax_interop"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.