Woro Feed
Hugging Face — AI news
The labs, the newsrooms, and WORO's take on what each story means for a small business. Updated every half hour.
AllWoro AIOpenAIGoogle AIGoogle DeepMindMeta AIMicrosoft AINVIDIAHugging FaceAWS Machine LearningThe VergeTechCrunchMIT Technology ReviewWIREDVentureBeat
Accelerating vision-language models with LFM2.5-VL-DSparkHow to Use NVIDIA Warp and MjWarp to Accelerate Robotics Simulation and Learning WorkflowsHow UK AISI and EvalEval Are Making Benchmark Results ReproducibleTransformers now runs llama.cpp quantsJun Kim, oMLX creator and maintainer, joins Hugging Face to support the MLX communityPruning LLMs Like a Physicist: Block Removal as an Ising Optimization Problemtokenizers v1: encode, decode and scaling, measuredYour Agent Aced the Task. Will It Do It Again?Async GRPO with LoRA across HF Jobs: a bucket, a proxy, and no NCCLRebuilding AUTOMATIC1111 with Gradio WorkflowNeoMME: an efficient Multimodal-native and Multilingual EncoderFine-tuning a 350M Model for Better Structured Outputs in 100 GRPO StepsGive Your Coding Agents a Memory You OwnTraining a coding model to paint watercolours with TRL and OpenEnvBenchMIRT: What are LLM benchmarks actually measuring?Introducing @huggingface/kernels: 200+ WebGPU Kernels for Local AIThe Open ASR Leaderboard Adds Its First Global South LanguageTraining and Finetuning Multi-Vector Embedding Models with Sentence TransformersGranite 4.2 LLMs: How They're BuiltQuantization-Aware Healing: a compressed, 4-bit model that outperforms its full-precision original