OpenXLA

A community-driven, open source ML compiler ecosystem, using the best of XLA & MLIR.

[Google Open Source @GoogleOSS]

🛠 Fast or slow? Build a predictor for a chance to win a prize!

Google is hosting a @kaggle competition to predict the performance of computation graphs from popular open-source AI models on TPUs.

Submit your results by November 17 to enter → https://t.co/EI0dna2mLG

[PyTorch @PyTorch]

With PyTorch + OpenXLA coming together, we're excited about the path forward to create an open stack for large scale AI development: https://t.co/9aLI7zDlsd

Including:
- Training large models
- Optimized model deployment
- Ecosystem integration with Lightning, Ray & Hugging Face

[PyTorch @PyTorch]

With the launch of 2.0, we’re happy to share the latest on PyTorch + XLA!

This version includes:
📊 Experimental support for TorchDynamo
💾 PJRT as the default runtime
🤗 Support for FSDP and GSPMD with integration in Hugging Face Transformers

https://t.co/US9ZU4Gu2d @GoogleOSS

[OpenXLA @OpenXLA]

Come see #OpenXLA at #GTC23 – Today at 10AM PDT → https://t.co/43AELXG4Lh

Learn how OpenXLA boosts #ML performance and development speed from @davidmajnemer @herhut @eugene_burmako

[Cerebras @cerebras]

“OpenXLA helps extend our user reach and accelerated time to solution by providing the Cerebras Wafer-Scale Engine with a common interface to higher level ML frameworks,” says our VP of Product, Andy Hock

Learn more about the OpenXLA Project here: https://t.co/qFYNOOEElE

[Alibaba Cloud @alibaba_cloud]

Looking forward to cooperating with #OpenXLA to develop an advanced ML compiler that delivers superior performance and user experience for #AlibabaCloud customers.

[Arm @Arm]

📢 You can now use and contribute to the #OpenXLA project!

This industry-wide collaboration is simplifying ML software development, making it easier to compile and optimize models from all leading ML frameworks.

Get started today on @GitHub.

[Hugging Face @huggingface]

You can now use and contribute to #OpenXLA 🚀🤗

Making it easy to run any model efficiently on any hardware is a deep technical challenge, and an important goal for our mission to democratize good ML!

[AMD @AMD]

AMD is committed to support open source software and are excited to partner with @Google on #OpenXLA. Learn how we are working with OpenXLA to support cross platform portability for #ML models across CPUs, GPUs, FPGAs and more.

[Jeff Dean @JeffDean]

Open source software gives everyone the opportunity to help create breakthroughs in AI. At Google, we’re excited to collaborate on the OpenXLA Project to foster great AI tooling (great ML performance, framework compatibility, highly configurable & more).

https://t.co/ea4nbH1sNg

[PyTorch @PyTorch]

#OpenXLA gives developers significant speed-ups using XLA for PyTorch models, and we're proud to be a part of this project!

[Cerebras @cerebras]

Cerebras builds leading AI systems and software to make large-scale AI easy and accessible to organizations. We are proud to see OpenXLA, a solution that makes frameworks easier to deploy across different hardware options, being made generally available and accessible to all.

[Software Dev Tools @IntelDevTools]

.@Intel is excited to partner with @GoogleOSS and other #OpenXLA stakeholders to allow #ML frameworks to leverage advanced AI architectures. We are committed to democratizing access to #MachineLearning through our AI-optimized hardware and #oneAPI powered software.
#openecosystem

[Gadi Hutt @gadi_hutt]

Say hi to #OpenXLA! At AWS we are very excited to be part of the founding team of #OpenXla , enabling customers with access to performant, scalable, and extensible AI infrastructure, driven by an open source community. AWS Trainium…https://t.co/s6WthPrhgl https://t.co/1vQYIGwHxE

[Graphcore @graphcoreai]

Graphcore is proud to support #OpenXLA - continuing our commitment to open source in AI and making it easier to move ML workloads between accelerators, including the IPU.

[NVIDIA AI Developer @NVIDIAAIDev]

We are happy to see the #ML developer community get behind #OpenXLA to take advantage of state-of-the-art #AI infrastructure.

[Google Open Source @GoogleOSS]

You can now contribute to #OpenXLA on GitHub → https://t.co/GELyVwlRd9

Speed up, scale, and execute your #ML models on different accelerators. A collaborative OSS effort among industry leaders to address #AI infrastructure fragmentation.

[Google Cloud @googlecloud]

We’re committed to open ecosystems of all kinds, and this extends to AI/ML—we believe no single company should own AI/ML innovation.

Read how we're taking this commitment a step further as a founding member and contributor to the OpenXLA Project ↓
https://t.co/9E7fOirxhv

[Jim St. Leger @JimStLeger]

.@Intel is working with @Google on OpenXLA to ensure there is an open, multi-platform approach to AI/ML. #IntelON #IAmIntel #AI #ML #OpenXLA

[Mehdi Amini @JokerEph]

#C4ML (compilers for ML) workshop today in Montreal!
https://t.co/OoRBlstZkd

After an overview of TorchInductor, @Mahesh0112 talked about #IREE / #OpenXLA ! Very interesting dive into the #MLIR based codegen pipeline and the runtime abstractions exposed by IREE.