🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

After 6 months of hard work, happy to share JAXopt: hardware accelerated, batchable and differentiable optimizers in JAX github.com/google/jaxopt We ambition to cover many use cases in ML: stochastic optim of DL models, constrained/non-smooth optim, bi-level optim, optim layers...

Today, I’m excited to announce the release of Levanter 1.0, our new JAX-based framework for training foundation models, which we’ve been working on Center for Research on Foundation Models. Levanter is designed to be legible, scalable and reproducible. crfm.stanford.edu/2023/06/16/lev…

Built with JAX!

Excited to share Penzai, a JAX research toolkit from Google DeepMind for building, editing, and visualizing neural networks! Penzai makes it easy to see model internals and lets you inject custom logic anywhere. Check it out on GitHub: github.com/google-deepmin…

Many of you are excited about H100 attention, so it’s a good time to show you Mosaic GPU: a Python DSL for H100s. The attention example matches FA3 performance, while being only ~200 lines of Python: github.com/google/jax/blo… It's easy to install too! Latest JAX packages have it.

Finally got around to writing a guide for matrix multiplication on TPUs using Pallas. Check it out! jax.readthedocs.io/en/latest/pall…

I've finally landed my first proper JAX feature since joining the team: a supported "foreign function interface", which makes it easier to call into external libraries from within JAX code. Check it out: jax.readthedocs.io/en/latest/ffi.…

Modula x JAX = Modulax theseriousadult is cracked and ported Modula into JAX in a few days. I haven't had a chance to test yet, but I'm really excited about this project. Tagging Roy Frostig and Matthew Johnson github.com/GallagherComma… (1/3)

We now have a guide to writing distributed communication on TPU using Pallas, written by Justin Fu! jax.readthedocs.io/en/latest/pall… Overlapping comms + compute is a crucial performance optimization for large scale ML. Write your own custom overlapped kernels in Python!

Our online book on systems principles of LLM scaling is live. We hope that it helps you make the most of your computing resources. Enjoy!

Training our most capable Gemini models relies heavily on our JAX software stack + Google's TPU hardware platforms. If you want to learn more, see this awesome book "How to Scale Your Model": jax-ml.github.io/scaling-book/ It was put together by my Google DeepMind colleagues

A nice and concise R1 inference jax:tpu port by rdyro. Good for both reading and running. Watch the repo for more.

Llama 4 inference in pure JAX! Expert/tensor parallelism with int8 quantization. Contributions welcome!