The name is the button.

An agent writes the code now, and a human thumb hovers over Accept all a hundred times a day. The honest question of this decade is what, exactly, you just accepted. So I read the diff.

Eight years as a backend engineer in Stockholm, back in Tokyo since 2023, and I still ship something small most nights, usually badly. My one rule: the only real review of a model is the thing you built with it. That is why there is no leaderboard on this site.

Around a hundred and forty sources come through — the labs' own posts, the papers, the tools, the runtimes and languages nobody is watching yet, the chip under the demo, and the newsletters people actually forward. What is moving gets counted so the week reads at a glance, and every few days I explain one mechanism in plain words.

A story here is a headline, a short excerpt, one image and the link to whoever wrote it. We never reproduce the post; the images belong to their authors. A post written in Japanese appears in Japanese, exactly as it was published, and the link goes to the original.

— Yuki Halvorsen, Tokyo

← Accept All Archive

PyTorch BlogTools

Hardware-Agnostic Models in vLLM

September 22

TL;DR To achieve state-of-the-art performance at the frontier, vLLM is changing its internal implementation in ways that make it incompatible with fullgraph torch.compile.

Open weights

Read at PyTorch Blog ↗

One letter a week, in your inbox.

Latest

Hardware-Agnostic Models in vLLM

Related

ToolsHow Shopify built a continual learning loop with PyTorch and vLLM PyTorch Blog

ToolsHelion x 🤗 HF Kernels: Building and Shipping Out-of-the-box Performant Kernels PyTorch Blog

ToolsPyTorch Conference China 2026: Advancing the Open Source AI Stack PyTorch Blog

SiliconNVIDIA Isaac ROS 5.0 Advances Agentic, Open Source Robotics Development NVIDIA Blog

ResearchIntrinsic open sources key parts of its platform for easier development The Robot Report

AsiaShanghai AI Lab Ships Intern-S2-397B Science Multimodal Model With Memory Decoder Pandaily