A machine learning compiler for GPUs, CPUs, and ML accelerators

Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more

A simple, performant and scalable Jax LLM!

Orbax provides common checkpointing and persistence utilities for JAX users

Source code to pbrt, the ray tracer described in the forthcoming 4th edition of the "Physically Based Rendering: From Theory to Implementation" book.

MLCommons Algorithmic Efficiency is a benchmark and competition measuring neural network training speedups due to algorithmic improvements in both training algorithms and models.

Dopamine is a research framework for fast prototyping of reinforcement learning algorithms.

CartPole-v0 via PPO with GAE, PyTorch