Jeff Boudier 🤗

Reachy Mini was everywhere at GTC 😍 Why? Because it’s the perfect interface to interact with your local, private agents (hey, Nemoclaw!) Thanks to our partners @nvidia , @Dell, @ASUS , @Lenovo, @HP and @msitweets for showing the power of private agents running locally on GB10. So much fun to be able to talk to them, showing them what you want instead of reaching for an app and typing, uploading images. 🦞+ 🤖 = Reachy Pinchy?

🕺 Kimodo v1 - Generate body movement from prompts huggingface.co/spaces/nvidia/…

🚕 Alpamayo 1.5 - Autonomous Vehicle Model and Datasets huggingface.co/blog/drmapavon…

💚🤗💚 Jensen showing @huggingface during GTC keynote, where @NVIDIAAI dropped amazing new open models, datasets and blogs! Some of my favorites, links in comments: 🧠 Nemotron 3 Super 120A12B - Reasoning LLM 🏥 Open-H-Embodiment - Healthcare Robotics Dataset 🩻 Cosmos-H-Surgical-Simulator - World Foundation Model 🚕 Alpamayo 1.5 - Autonomous Vehicle Model and Datasets 🕺 Kimodo v1 - Generate body movement from prompts! And SO MUCH MORE to explore at huggingface.co/nvidia/collect…

So you want to deploy the new @nvidia Nemotron 3 Super... but what's the optimal config to max out throughput or latency? 🧐 We ran the numbers as part of the Goodput Scenarios analysis 📊 on @Dell Enterprise Hub 👀 dell.huggingface.co/authenticated/…

Hugging Face Builders is a global community program that puts local leaders at the center of the open-source AI movement 🤗 If you're passionate about open AI and love bringing people together, this is your invitation to lead ✉️ Apply for to build the Paris chapter today ➡️

Bitsandbytes now has full ROCm/CUDA feature parity! 🧵 See thread for more details.

@mervenoyann @NVIDIAAIDev def see a Spark in your eye! now you gotta Reachy for the stars ✨

what else can a girl ask for, @NVIDIAAIDev has sent me a DGX Spark for my upcoming birthday 💚 no joke I cried out of happiness 🥹 I will use it to label datasets and later train models and open-source them, but first will play with Nemotron-3 Super 🤗

Come work with @mervenoyann, Alvaro and I! 🤗🤗🤗

How to fine-tune with TRL: github.com/huggingface/tr…

All the weights: huggingface.co/collections/nv…

🤯 @NVIDIAAI just dropped Nemotron 3 Super on @huggingface - and it's ready to deploy on @Dell 🔥 120B total / 12B active — hybrid Mamba MoE, FP4 native 🔓 Super Open: weights + data + recipe, all open 🤖 Super Agentic: 1M context, 37 RL envs, built for real agentic tasks ⚡ Super Efficient: 2.2× faster than GPT-OSS-120B, 7.5× vs Qwen3.5 in FP4 Run with Transformers, fine-tune with TRL, deploy on-prem with Dell Enterprise Hub 👇 💚💛💙 Great collab @ctnzr @MichaelDell @ClementDelangue

@jackadoresai @huggingface thanks! Latency will benefit from Xet chunk-defined deduplication, plus with cloud prewarming you can set the storage close to where your training pipelines run

@jeffboudier @huggingface 3x cheaper than S3 is wild. Curious about latency trade-offs though. Could really change training pipelines if performance holds up.

What does AI-Native Storage look like? 🔮🧐 Our answer: HF Storage Buckets 🪣🪣🪣 a new Repo Type on @huggingface. Accelerate your training runs. Power your agents. Reduce your AI infrastructure costs. Starting at $8/TB/mo — that's 3x cheaper than S3.

Check it out → huggingface.co/blog/storage-b…

Hugging Face just launched buckets 🪣 ◆ Starting at 12TB per month (with volume discounts) ◆ CDN Pre-Warming for Faster Downloads ◆ XET Deduplication to help with speed + cost Best for large volume AI workloads!