JAX

Python library for composable transformations of NumPy-style code: autodiff, JIT compilation and vectorization.

by Google No reviews yet Open source Free jax-v0.11.2 Updated 2 days ago 69 88.0 =

Visit website Source code Compare with Transformers

Overview

Python library for composable transformations of NumPy-style code: autodiff, JIT compilation and vectorization.

Updated 2 days ago

10 releases in the last 12 months.

Ratings appear once verified reviews arrive. Until then the rank uses public signals only: release freshness and repository adoption.

See all alternatives

Transformers 4 93.5 =

Python library with model definitions and pretrained weights for text, vision and audio transformers.

PyTorch 17 91.8 =

Python deep learning framework with tensors, automatic differentiation and GPU acceleration.

scikit-learn 33 90.2 =

Python library for classical machine learning: classification, regression, clustering and preprocessing.

TensorFlow 153 69.2 =

Machine learning platform for building, training and deploying models on servers, browsers and devices.

Rank receipt: JAX

#69 overall · #4 in category

88.0/100

Data from GitHub Last sync 2026-09-18 Computed on 2026-09-19 Sponsorship weight 0%, never counted Read the full formula

GitHub Synced https://github.com/jax-ml/jax Last sync 2026-09-18

The vendor's own feeds. Contact details inside them are discarded, never stored.