Keras 3 is a deep learning framework that runs on JAX, TensorFlow, PyTorch, or OpenVINO for inference.
64.4K19.8KPythonApache-2.0Updated yesterday
unifyai/ivy
A Python library that converts machine learning code between frameworks such as PyTorch, TensorFlow, JAX and NumPy via ivy.transpile.
- Stars
- 14,169
- Forks
- 5,443
- Open issues
- 988
- Last commit
- 7 hours ago
What is it?
Ivy converts machine learning models, tools and libraries from one framework to another. Its main entry point, ivy.transpile, takes a function, class or module written for one framework and returns an equivalent version that runs on a target framework, so code written for PyTorch can be used from TensorFlow without rewriting it by hand.
Transpilation can happen eagerly when a function or class is passed in, or lazily when a whole module is passed in, in which case functions are converted the first time they are called. Ivy also provides ivy.trace_graph, which traces a function into a functional computational graph for a chosen target framework.
The README lists PyTorch as a supported source and TensorFlow, JAX and NumPy as supported targets, with other source combinations marked as planned.
Key features
- 01Converts functions, classes and whole modules between ML frameworks with ivy.transpile.
- 02Supports lazy transpilation of entire libraries such as kornia.
- 03Traces functions into computational graphs with ivy.trace_graph.
- 04Installable with pip.
Use cases
- Reusing PyTorch model code inside a TensorFlow project.
- Running a PyTorch-based utility library on JAX or NumPy.
- Porting ML code between frameworks without manual rewrites.
Topics
FAQ
- What is ivy?
- A Python library that converts machine learning code between frameworks such as PyTorch, TensorFlow, JAX and NumPy via ivy.transpile. Ivy converts machine learning models, tools and libraries from one framework to another. Its main entry point, ivy.transpile, takes a function, class or module written for one framework and returns an equivalent version that runs on a target framework, so code written for PyTorch can be used from TensorFlow without rewriting it by hand.
- Is ivy free to use? What license does it have?
- ivy is open source under the Other license. Check the license terms before using it in commercial products.
- Is ivy still maintained?
- Actively maintained. The last commit to ivy was on Oct 22, 2026.
- What can I use ivy for?
- Reusing PyTorch model code inside a TensorFlow project. Running a PyTorch-based utility library on JAX or NumPy. Porting ML code between frameworks without manual rewrites.
- Where can I get ivy?
- The source code and installation instructions are on GitHub: https://github.com/unifyai/ivy
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
Viewer for neural network, deep learning and machine learning models, supporting ONNX, TensorFlow, PyTorch, Core ML and many other formats.
33.6K3.2KJavaScriptMITUpdated 4 days ago
A cross-platform accelerator for running ML models exported to ONNX format, with training support for PyTorch scripts.
22.1K4.3KC++MITUpdated 1 hour ago
Python library and hosted platform for tracking, visualizing and managing machine learning experiments, datasets and models.
11.3K907PythonMITUpdated 2 hours ago
A 12-week, 24-lesson beginner AI curriculum from Microsoft covering symbolic AI, neural networks, computer vision, NLP and ethics, with notebooks in PyTorch and TensorFlow.
Learning Resources#artificial-intelligence#machine-learning#deep-learning#neural-networks#computer-vision
70.1K13.6KJupyter NotebookMITUpdated last month
A Python deep learning framework offering GPU-accelerated tensor computation and dynamic neural networks built on autograd.
104.9K32.4KPythonOtherUpdated yesterday