Please welcome MegaBlocks to the Databricks family! databricks.com/blog/bringing-…

Llama3 8B and 70B are out, with pretty exciting results! * The ~400B is still training but results already look promising. * Meta's own Chat interface is also live at meta.ai * TorchTune integration is shortly going live: github.com/pytorch/torcht…

Last week, I spoke about AI and regulations at an event at the U.S. Capitol attended by legislative and business leaders. I’m encouraged by the progress the open source community has made fending off regulations that would have stifled innovation. But opponents of open source are

Very nice results from a Google AI research effort on a general purpose time-series prediction model that gives good zero-shot performance to new forecasting tasks.

🚀 PyTorch TPU proud moment: vLLM TPU 🚀 At Google Cloud Next '25 keynote, Amin Vahdat announced vLLM on TPU. This is a HUGE breakthrough moment for TPU adoption! 🔥 Thank you Woosuk Kwon for your awesome contribution and collaboration.

.Fei-Fei Li on why LLMs will struggle to solve spatial intelligence: "Language is fundamentally a purely generated signal." "You don't go out in nature and there's words written in the sky for you." "There is a 3D world out there that follows laws of physics... to fundamentally

Today Thinking Machines Lab is launching our research blog, Connectionism. Our first blog post is “Defeating Nondeterminism in LLM Inference” We believe that science is better when shared. Connectionism will cover topics as varied as our research is: from kernel numerics to

BREAKING 🚨: Meta announced Meta Ray-Ban Display AI Glasses with an EMG Wristband! Did Zuck just kill the phone industry? 👀 Honestly, a wristband is a HUGE enabler, but there are significant questions about its quality.

Sharing our second Connectionism research post on Modular Manifolds, a mathematical approach to refining training at each layer of the neural network

GPUs are expensive and setting up the infrastructure to make GPUs work for you properly is complex, making experimentation on cutting-edge models challenging for researchers and ML practitioners. Providing high quality research tooling is one of the most effective ways to

it is common knowledge but i absolutely love how torch compile is able to fuse separate function calls doing some elementwise ops into a single triton kernel. it's like witnessing magic. we don't appreciate torch compile enough.

Hanging out with Project Jupyter co-founder Brian Granger. If not for him and fernandoperez.org we wouldn’t have the coding notebooks we use daily in AI and Data Science. Very grateful to him and the whole Jupyter team for this wonderful open-source work!

It’s an honor to have received the Queen Elizabeth Prize for Engineering along with my fellow laureates! But it’s also a responsibility. AI’s impact to humanity is in the hands of all of us.

2025 was the year when artificial intelligence’s full potential roared into view, and when it became clear that there will be no turning back. For delivering the age of thinking machines, for wowing and worrying humanity, for transforming the present and transcending the

This came as a total surprise this morning. Very humbled… 🙏 AI is built by generations of technologists, starting with the daring question of “can machines think?” by Alan Turing. It will be further developed, used and governed by many and all of us! Let’s keep our AI mission

Today’s "Inside Helion Live Q&A" brought the #PyTorch community together with Jason Ansel, Oguz Ulgen, Wei (Will) Feng, and Jongsok Choi from Meta’s PyTorch Compiler and Helion teams. The discussion explored how Helion approaches kernel authoring, #AIInfrastructure performance, and

Zhipeng (Jason) Wang, PhD (Zhipeng Wang 🇺🇦) explains how DeepSpeed supports ML training research and why joining PyTorch Foundation benefits researchers and developers working on AI training workloads. 🔗youtu.be/67719mlOSp0 #PyTorch #DeepSpeed #OpenSourceAI #AIInfrastructure

Three days ago I left autoresearch tuning nanochat for ~2 days on depth=12 model. It found ~20 changes that improved the validation loss. I tested these changes yesterday and all of them were additive and transferred to larger (depth=24) models. Stacking up all of these changes,