Hugging Face

The last 8 weeks

31
postsFirst-partypeak 6
3
outlets reportedOn 3 postspeak 1
No. 26
by popularityDown 1 since last week

The last 8 weeks

31
postsFirst-partypeak 6
3
outlets reportedOn 3 postspeak 1
No. 26
by popularityDown 1 since last week

@huggingface RSS

Hugging Face· 1 day ago

Welcome Beam: a 501B open model you can own, customize, and scale

Reflection's 501B MoE model is open under Apache 2.0

New model

Hugging Face

American machine learning company

The AI community building the future.

Founded: 2016
Based: New York City
Site: huggingface.co

@huggingface on X Wikipedia ↗

Everything else, by date

Granite Switch 4.1: Chat, Decisions, and Speech from One EndpointModel update15 Oct
Apertus 1.5 Release NotesModel update15 Oct
Welcome Muse Spark Open: open-weight frontier multimodal mixture-of-experts model from MetaOpen-weight release15 Oct
Welcome FLUX.3 Image Klein in DiffusersOpen-weight release15 Oct
The model that didn't exist, so you made it yourself8 Oct
AutoSynthData: Generating Training Data for Enterprise AgentsResearch paper2 Oct
Open TTS Leaderboard: Scalable Evaluation for Multilingual Text-to-Speech and Voice Cloning30 Sep
Getting the Source Right, Not Just the Fact: Source-Aware Verification for MCP AgentsResearch paper29 Sep
Welcome RL Environments to the hubCapability change28 Sep
How UK AISI and EvalEval Are Making Benchmark Results Reproducible22 Sep
Transformers now runs llama.cpp quantsCapability change22 Sep
Jun Kim, oMLX creator and maintainer, joins Hugging Face to support the MLX community22 Sep

+37 more posts

Pruning LLMs Like a Physicist: Block Removal as an Ising Optimization ProblemResearch paper21 Sep
tokenizers v1: encode, decode and scaling, measuredDeveloper tool21 Sep
Your Agent Aced the Task. Will It Do It Again?Developer tool15 Sep

Who reports on Hugging Face

Stories each covered · of 3 covered

Rohan Paul1
Scan1
Unite1

Hugging Face

American machine learning company

The AI community building the future.

Founded: 2016
Based: New York City
Site: huggingface.co

@huggingface on X Wikipedia ↗

Who reports on Hugging Face

Stories each covered · of 3 covered

Rohan Paul1
Scan1
Unite1

Async GRPO with LoRA across HF Jobs: a bucket, a proxy, and no NCCLDeveloper tool10 Sep

Rebuilding AUTOMATIC1111 with Gradio WorkflowDeveloper tool10 Sep

IBM releases SOTA Granite Time Series PatchTST-FM-r2 model with commercial-friendly licenseNew model9 Sep

Safety for Whom? Refusing the Right Subset of a Topic, Not the Whole TopicResearch paper8 Sep

Fine-tuning a 350M Model for Better Structured Outputs in 100 GRPO StepsDeveloper tool3 Sep

Give Your Coding Agents a Memory You OwnDeveloper tool3 Sep

Training a coding model to paint watercolours with TRL and OpenEnv3 Sep

Real-Time Intelligence with IBM Time Series Models on ConfluentPartnership2 Sep

Introducing @huggingface/kernels: 200+ WebGPU Kernels for Local AIDeveloper tool1 Sep

The Open ASR Leaderboard Adds Its First Global South LanguageCapability change28 Aug

Training and Finetuning Multi-Vector Embedding Models with Sentence TransformersDeveloper tool26 Aug

Granite 4.2 LLMs: How They're BuiltNew model25 Aug

Extremely Fast and Accurate Transcription with Granite Speech 5.0 Turbo CTCNew model25 Aug

Quantization-Aware Healing: a compressed, 4-bit model that outperforms its full-precision originalResearch paper25 Aug

Wire It, Run It, Deploy It: AI Workflows in GradioCapability change25 Aug

How Hugging Face Inference Endpoints, Jobs, and Buckets Power Search on Papers with Code21 Aug

Measuring benchmark optimization in speech recognitionProduct launch21 Aug

How Much Memory Does Your Agent Actually Need?Research paper18 Aug

Multi-Vector (Late Interaction) Embedding Models with Sentence TransformersModel update18 Aug

Same Cluster, 33 Points More Utilization: What Changed Was the Order17 Aug

State of Open Models: Summer 2026 ObservationsResearch paper14 Aug

What We Learned by Reproducing 2,200 papers from ICMLResearch paper13 Aug

Thinking of ACE? We Can Do It with Fewer TokensResearch paper11 Aug

Making Knowledge Distillation Cheap Enough to Run at ScaleResearch paper10 Aug

Meta is back with Muse Glimmer: local, agentic, multimodal, and open sourceNew model10 Aug

Baseten on Hugging Face Inference Providers 🔥Infrastructure6 Aug

GPU Management: Why Idle GPUs Are the New Grounded Aircraft30 Jul

Anatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 IncidentResearch paper27 Jul

Bringing Nunchaku 4-bit Diffusion Inference to DiffusersDeveloper tool23 Jul

Grabette: an open system to record robot-manipulation dataDeveloper tool21 Jul

Newer Models, Same Advantage16 Jul

Security incident disclosure — July 202616 Jul

Model Routing Is Simple. Until It Isn’t.15 Jul

Introducing Real World VoiceEQ: Measuring the human quality of voice AIResearch paper15 Jul

The last 8 weeks

Lead story

The last 8 weeks

Lead story

Hugging Face

The last 8 weeks

Welcome Beam: a 501B open model you can own, customize, and scale

Hugging Face

Who reports on Hugging Face

Hugging Face

Who reports on Hugging Face