I gave a talk on "Build-to-own AI: Agentic Development for Humans" at PyCon Armenia: how coding agents change the way we build ML systems and why code and programmable interfaces matter more than ever, not less. Here's the whole thing as a thread 🧵 (1/18)

You can think of human-agent collaboration like a centaur, and we imagine we're the head. But if the computer tells you what to do and you just click buttons and report back, you've got the horse's head and the human butt. That's why so many AI workflows feel bad. We need shared interfaces.

People not in tech have been interested in what happened with the Open AI agent/Hugging Face hack. So, I "Explain it Like I'm 5"-ed it and made a cartoon in 8 panels. Apologies for some anthro-ing. 🫤 m-mitchell.com/HF-hack-cart... 1/8

Volkswagen blocks custom ROM users from VW app https://www.heise.de/en/news/Volkswagen-blocks-custom-ROM-users-from-VW-app-11370557.html?utm_source=flipboard&utm_medium=activitypub Posted into heise online international @heise-online-international-heiseonline

Volkswagen blocks custom ROM users from VW app

Volkswagen has blocked access to its app for users of custom ROMs like GrapheneOS or LineageOS. The reason is the Google Play Integrity API.

heise.de

Moved from the Bluesky PDS to Eurosky. The migration was really easy and I think everything was migrated without issues 🤞🏻. @eurosky.social is hosted in Europe by a Dutch non-profit foundation.

We recently made a lot of updates to the kernels ecosystem, from improving security, adding a dedicated kernels repo type on the hub, to support for Apache tvm-ffi. We made a write-up of these recent changes in this blog post: huggingface.co/blog/revampe...

🤗 Kernels: Major Updates

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

huggingface.co

We have recently added Torch Stable ABI support to @hf.co kernels and kernel-builder. So you can now target a specific minimum Torch version for long-term support. We're rolling out Stable ABI builds of kernels in kernels-community, starting with Flash Attention 3: huggingface.co/kernels/kern...

kernels-community/flash-attn3 at v1

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

huggingface.co

I switched from Claude Sonnet/Opus to mostly using Kimi K2.7 Code and GLM-5.2, and... nothing bad happened. The @hf.co Inference API makes it super easy to switch between different models and providers: huggingface.co/inference/mo... Open models are really competitive - great pricing and fast.

Hugging Face Inference Providers · Supported Models

Exhaustive list of models supported by Hugging Face Inference Providers with detailed features and performance metrics.

huggingface.co

Looking for a software engineer to help maintain your OS repo? I currently have availability for a part-time open-source engagement. Preferably, focused on data, evaluations, NLP/ML/LLM. I work as a consultant through my company OxyKodit. DM me if you'd like to explore working together!

The flexibility of #nix and #nixos is insane. What if you are building compute kernels with Nix, but you need them to be compatible with manylinux_2_28 (based on Enterprise Linux 8)? Simple, repackage the EL8 gcc toolset RPMs as Nix derivations and wrap it in a stdenv. github.com/huggingface/...

kernels/nix-builder/pkgs/manylinux at main · huggingface/kernels

Build compute kernels and load them from the Hub. Contribute to huggingface/kernels development by creating an account on GitHub.

github.com

Messi xkcd.com/3260/

Here's an easter egg in the new Lego Batman that I think all of you will REALLY appreciate. It's so good, I had to make a video.

We released kernels 0.15.1, packed with new features, including: * Torch stable ABI support * Better manylinux_2_28 support * A skill for making XPU kernels * Better offline support * Docs on IDE support for local development. github.com/huggingface/...

Want to be a beta partner of our upcoming product for agentic NLP? We're looking for teams who need help with their NLP projects, especially around tasks like information extraction. Ellf is a virtual assistant for NLP projects – think Claude Code, but with NLP superpowers, or "Vibe NLP" ✨

Safetensors has become one of the cornerstones of machine learning tooling for safe loading of arrays. Safetensors has now joined the @pytorch.org Foundation: pytorch.org/blog/pytorch... The repository has also moved to a new home: github.com/safetensors/...

Happy Trans Day of Visibility! 🏳️‍⚧️ The mere act of existing and being visible shouldn’t be as fraught as it is today. Let’s keep fighting to set things right. en.wikipedia.org/wiki/Interna...

International Transgender Day of Visibility - Wikipedia

en.wikipedia.org

I got tables working nice

More embedding models and an even more reliable inference engine is what you get with @hf.co Text Embeddings Inference v1.9.0 💥 More in the thread 🧵

Kernels now has an agent skill to write custom Hub kernels: huggingface.co/blog/custom-... Awesome work by @benburtenshaw.bsky.social and Sayak Paul! 🔥

Custom Kernels for All from Codex and Claude

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

huggingface.co

kernels 0.12 is out! 🎉 Changes: * Support for kernel version branches to gracefully roll out kernel API changes. * Support for PyTorch 2.10. * kernel-builder is now merged into the kernels repo. * Initial support for standardized kernel benchmarks. github.com/huggingface/...

Release v0.12.0 · huggingface/kernels

New features Merge of kernels and kernel-builder repositories kernel-builder has been merged into the kernels repository. This makes it easier for us to coordinate changes that affect both the kern...

github.com

DeepSeek R1 dropped one year ago 🐳 and a lot has changed. With Irene Solaiman, we’re launching a blog series on @hf.co about how that moment reshaped AI + open source in 2025, starting with strategic shifts and the explosion of new open models in China! huggingface.co/blog/hugging...

One Year Since the “DeepSeek Moment”

A Blog post by Hugging Face on Hugging Face

huggingface.co

🔥I am super excited for the official release of an open-source library we've been working on for about a year! 🪄interpreto is an interpretability toolbox for HF language models🤗. In both generation and classification! Why do you need it, and for what? 1/8 (links at the end)

👀 Look what 🎅 has broght just before Christmas 🎁: a brand new Research Master in Natural Language Processing at @facultyofartsug.bsky.social @rug.nl Program: www.rug.nl/masters/natu... Applications (2026/2027) are open! Come and study with us (you will also learn why we have a 🐮 in our logo)

Natural Language Processing

How do you build Large Language Models? How do humans experience Natural Language Processing (NLP) applications in their daily lives? And how can we...

rug.nl

I finally made a page on my Dish activation function, replacing my deleted Tweet: danieldk.eu/Dish-Activat... It's a non-monotonic function similar to GELU/SiLU, but does not require elementary functions, making it faster on various hardware. I'll leave the empirical evaluation to someone else 😁.

Training LLMs end to end is hard. But way more people should, and will, be doing it in the future. The @hf.co Research team is excited to share their new e-book that covers the full pipeline: · pre-training, · post-training, · infra. 200+ pages of what worked and what didn’t. ⤵️

The Hub is on 100% on Xet. 🚀 A little over a year ago, @hf.co acquired XetHub to unlock the next phase of growth in models and datasets. huggingface.co/blog/xethub-... In April, there were 1,000 Hugging Face repos on Xet. Now every repo (over 6M) on the Hub is on Xet.

We made a blog post on how you can use kernel-builder to develop and build compute kernels for the @hf.co Kernel Hub: huggingface.co/blog/kernel-...

From Zero to GPU: A Guide to Building and Scaling Production-Ready CUDA Kernels

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

huggingface.co

Yesterday we released support for GPT OSS (the new OpenAI open weight model) across the @hf.co ecosystem. The latest Transformers now integrates support for the kernels package and uses kernels from the HF Kernel Hub to run models like GPT OSS as fast as possible. 🚀 huggingface.co/blog/welcome...

Welcome GPT OSS, the new open-source model family from OpenAI!

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

huggingface.co

Transformers 4.54.0 is out! This release adds support for compute kernels hosted on the Hub. When enabled, transformers can replace PyTorch layer implementations by fast, specialized kernels from the hub. github.com/huggingface/...

Release v4.54.0: Kernels, Transformers Serve, Ernie, Voxtral, LFM2, DeepSeek v2, ModernBERT Decoder... · huggingface/transformers

Important news! In order to become the source of truth, we recognize that we need to address two common and long-heard critiques about transformers: transformers is bloated transformers is slow O...

github.com