BlackHC/tfpyth

Putting TensorFlow back in PyTorch, back in TensorFlow (differentiable TensorFlow PyTorch adapters).

56
/ 100
Established

This library helps machine learning engineers or researchers integrate models built in TensorFlow with models built in PyTorch. It allows you to use a TensorFlow graph as a function within a PyTorch model, or vice-versa, making the entire combined system differentiable. This means you can train a single model that leverages components from both frameworks.

647 stars. No commits in the last 6 months. Available on PyPI.

Use this if you have existing model components or codebases in both TensorFlow and PyTorch and need to combine them into a single, end-to-end trainable system without rewriting either.

Not ideal if your entire workflow is confined to a single deep learning framework, or if you require direct GPU-to-GPU tensor transfers between frameworks (as data currently moves via CPU).

deep-learning-engineering model-integration machine-learning-research framework-interoperability neural-network-training
Stale 6m
Maintenance 0 / 25
Adoption 10 / 25
Maturity 25 / 25
Community 21 / 25

How are scores calculated?

Stars

647

Forks

96

Language

Python

License

MIT

Last pushed

Nov 30, 2020

Commits (30d)

0

Dependencies

2

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/BlackHC/tfpyth"

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