Hugging Face

Hugging Face (Twitter) RT @novita_labs: 🤗 DeepSeek V3.2 is live on @huggingface , supported by Novita Test it yourself👇

Hugging Face (Twitter) RT @lukas_m_ziegler: Something must change in robotics! 😖 As more billion-dollar players enter the space, more systems are closing off. APIs move behind paywalls, platforms lock down, and robots turn into black boxes. @pollenrobotics is taking the opposite route. With Reachy 2, everything is open: the code, the CAD files, the ROS 2 interfaces, the communication stack. If you want to swap the camera, redesign the gripper, or fork the whole system, you can, and that’s the point. It creates a culture where universities can run real experiments without licensing drama, and startups can prototype without waiting for vendor permission. Clone it on GitHub. Add behaviors. Share datasets. Fork the models. Push improvements upstream. The @huggingface influence is real. It’s two open ecosystems aligning around the same idea: robotics should be built by more than a handful of companies. It was so cool to meet the team behind Reachy 2 in Bordeaux!... Перейти на оригинальный пост

Hugging Face (Twitter) RT @cocktailpeanut: One great thing about Gradio is, it is API native. When you see a Gradio app, it always has an API. And now with Pinokio's "Ask AI" feature, you can instantly ask any AI agent--Claude code, Codex CLI, Gemini CLI, or even Cursor--how to use the API, and even some examples! https://twitter.com/cocktailpeanut/status/1995521657482399810#m

Hugging Face (Twitter) RT @ti_morse: My first interview with @ClementDelangue, cofounder & CEO of @HuggingFace. 0:52 Creating an affordable robot 11:24 Incentives 14:20 Hiring generalists 20:24 Designing a robot for builders 25:13 Focus is saying no 31:08 Slowly compounding for years 36:30 Optimizing for trust 39:12 Sisyphus 45:13 Great companies are just living embodiments of their founders 48:46 Two-way doors 54:17 Figuring out what you love to do 1:01:12 Living with uncertainty 1:04:36 Setting the right expectations with AI products 1:09:15 Why open-source matters 1:21:07 Real trends vs fake trends 1:30:30 Pavel Durov & France 1:33:59 Why China is leading in open-source AI 1:38:32 XAI open-sourcing Grok 3 1:43:17 Trends in robotics

Hugging Face (Twitter) RT @arcee_ai: Introducing Trinity, the start of a new open-weight MoE family. Rolling out today today: Trinity-Mini (26B-A3B) Trinity-Nano-Preview (6B-A1B) Download on HuggingFace. Free for limited time on OpenRouter.

Hugging Face (Twitter) Today we release the transformers version 5 RC! 🤗 With this, we enable e2e interoperability with our friends in ecosystem, ease up adding new models and simplify the library 🙌🏻 Read our blog to learn more: https://huggingface.co/blog/transformers-v5

Hugging Face (Twitter) RT @Thom_Wolf: This is quite mindblowing https://twitter.com/LysandreJik/status/1995558230567878975#m

Hugging Face (Twitter) RT @LysandreJik: Transformers v5's first release candidate is out 🔥 The biggest release of my life. It's been five years since the last major (v4). From 20 architectures to 400, 20k daily downloads to 3 million. The release is huge, w/ tokenization (no slow tokenizers!), modeling & processing.

Hugging Face (Twitter) RT @ArtificialAnlys: Introducing the Artificial Analysis Openness Index: a standardized and independently assessed measure of AI model openness across availability and transparency Openness is not just the ability to download model weights. It is also licensing, data and methodology - we developed a framework underpinning the Artificial Analysis Openness Index to incorporate these elements. It allows developers, users, and labs to compare across all these aspects of openness on a standardized basis, and brings visibility to labs advancing the open AI ecosystem. A model with a score of 100 in Openness Index would be open weights and permissively licensed with full training code, pre-training data and post-training data released - allowing users to not just use the model but reproduce its training in full, or take inspiration from some or all of the model creator’s approach to build their own model. We have not yet awarded any models a score of 100!... Перейти на оригинальный пост

Hugging Face (Twitter) RT @1littlecoder: Deepseek casually drops a 685 Billion Parameter model on Hugging Face with MIT license! The higher variant performance on par with GPT-5 and Gemini 3.0 Pro! And yet, this model is open weights coming with a Technical Paper! https://twitter.com/1littlecoder/status/1995455474112622752#m

Hugging Face (Twitter) RT @lhoestq: Just merged: 100x Faster scans with @DataPolars of large scale datasets on @huggingface⚡ fineweb-2 (8,257 files) 379 API calls -> 19 API calls finepdfs-edu (429 files) 139 API calls -> 1 API call Looking forward to the release ! See PR #25521 on github

Hugging Face (Twitter) RT @1littlecoder: The new Deepseek is here! Gold-medal performance in the 2025 International Mathematical Olympiad (IMO) and International Olympiad in Informatics (IOI).