A sister team of mine in ML Infrastructure is hiring multiple senior level PMs. If interested please reach out and I can make sure the HM has your profile. Here is one of the job reqs: Cheers! lnkd.in/gCnfeCn8

"OpenAI’s chatbot offers paraphrases, whereas Google offers quotes. Which do we prefer?" Obviously I am biased but this is still a great article and, as someone who spent many years working on video compression, the analogy speaks to me.. lnkd.in/gVrqgkEy

It's exciting that so many of the most interesting LLM and Generative AI startups are building on Google with Anthropic, Cohere and many others.. Having personally been involved with Midjourney and working with this really cool tea…lnkd.in/g3xYchx4 lnkd.in/gm9-ZSX3

The more things change, the more things stay the same. Yes, we have a PyTorch team at Google and yes they are incredible. Check out the latest work on Torch Dynamo, FSDP and what we see as the path forward collaborating with thi…lnkd.in/dTfs_arN lnkd.in/dgUQyzZB

PyTorch & OpenXLA, where do we go from here? As we celebrate the release of OpenXLA, PyTorch 2.0, and PyTorch/XLA 2.0, it’s worth taking a step back and sharing where we see it all going in the short to medium term. With PyTorch a…lnkd.in/g5WP6pdK lnkd.in/gwDy-tNU

Looking forward to leading off the speakers at the PyTorch Summit this week in Vancouver!! I'll talk about the humble beginnings of the PyTorch Community including some of the hard decisions we made to make it what it is today as the leading platform for…lnkd.in/gxnvwURY

Open source is most certainly powering generative AI, but will this continue? Open source and open collaboration have been the hallmarks of AI since its origins (a big reason I'm here!) with the current wave of innovation in LLMs and generative AI being…lnkd.in/gGyqYiSU

I'm excited (and honored) to speak at Fully Connected in SF next month talking about PyTorch and how a community driven approach became the foundation of the Generative AI. If you are in the area let's meet up!! #ai #community #pytorch lnkd.in/gxcWJNtW

Resharing this writeup from my old Metamate Tanya. Evaluating large models is a fast moving space and platforms like Lastmileai are really interesting to follow. Definitely check them out.. lnkd.in/gavwMRgU

I am so proud of my former colleagues Douwe and Amanpreet for what they are building and excited as an angel to even be a small part of supporting their success. I recommend following along as this team will do great things.. lnkd.in/gA6uszid

Thank you to Patrick, Sayak and Merve for the great collaboration on this hackathon!! #collaboration We are already seeing some really cool projects coming out (including the QR code project) that are starting to trend.. lnkd.in/gqTH-x8V

New launch from LastMile AI! For those that don't know, they are a startup with an incredible DNA for AI developer tools that is aiming to democratize LLMs for the enterprise. I worked closely with these folks at Facebook when they were building the Ben…lnkd.in/gnYUVS43

Wow - that was fast! Stealth mode to famous in a matter of weeks. Pretty sure it’s all because of Amanpreet! 🔥🔥(j/k Douwe) lnkd.in/gUfFhy8R

Congrats to the team on such amazing results! Guillaume and Timothee - I hope this helps you on your journey :) lnkd.in/gyA2VPwR

Congrats to Douwe, Amanpreet and team! lnkd.in/geBzeCKx

The Llama Guard model is now available on Anyscale Endpoints. Get started here: docs.endpoints.anyscale.com Example:

We are excited to partner with Meta and launch the Llama Guard model on Anyscale Endpoints! Get started: docs.endpoints.anyscale.com

In addition to today's launch of AI at Meta's new Llama model (Llama Guard), Anyscale Endpoints also recently launched open embeddings (gte-large). Lots more in the coming weeks... docs.endpoints.anyscale.com

With the release of Meta Llama 3.1 we are putting out the evaluation data for anyone looking to replicate our evals: huggingface.co/meta-llama

Just pushed significant updates to the llama-toolchain and llama-agentic-system repositories on GitHub. The biggest change is the addition of "Llama Stack Distributions" – a single server combining all the Llama Stack APIs together to let you get started quickly.