hugging face - Tech News from Simmons Systems

Tech News from Simmons Systems

hugging face

A collection of 38 posts

Hugging Face Breached by Autonomous AI Agent in Unprecedented Attack

Hugging Face, a major open-source AI platform used by over 50,000 organizations, discloses a significant breach in which attackers leverage an autonomous AI agent to infiltrate its production infrastructure. The intruders exploit two code-execution vulnerabilities within the platform's data-processing pipeline by using a malicious dataset, allowing them

Comparing Supervised and Reinforcement Learning for Fine-Tuning Reasoning LLMs

A recent discussion on the Hugging Face forums raises an important question for AI practitioners: whether to use supervised learning or reinforcement learning when fine-tuning a reasoning-focused large language model. The conversation centers on which approach produces better results for models designed to handle complex reasoning tasks. Supervised fine-tuning involves

Microsoft Launches Phi 3.5 AI Models to Rival Google and OpenAI

Microsoft introduces three new Phi 3.5 AI models that outperform competitors on key benchmarks while remaining open source for commercial use.

Coqui and Hugging Face Launch Open-Access XTTS Voice AI Model

AI startup Coqui partners with Hugging Face to release XTTS, a generative voice AI model that clones voices from just three seconds of audio. The open-access model supports 13 languages and offers cross-language cloning capabilities.

Qwen3 Launches Open-Weight AI Models Rivaling Top Industry Competitors

The Qwen team releases Qwen3, a new family of open-weight large language models that achieve top-tier benchmark results in coding and math. The release includes efficient Mixture-of-Experts models and smaller dense models available under the Apache 2.0 license.

xAI Releases Massive 314 Billion Parameter Grok-1 Model

xAI makes its Grok-1 open-weights model available on Hugging Face, offering a massive 314 billion parameter language model under the Apache 2.0 license. Users need a multi-GPU setup to run the model locally.

stable diffusion

New Stable Diffusion Model Brings Fast AI Art to Consumer Computers

Stability AI and its collaborators announce Stable Diffusion, a groundbreaking text-to-image model that runs efficiently on consumer GPUs. The highly efficient system democratizes AI art generation by requiring less than 10 GB of video memory.

DeepSeek's New Experimental Model Slashes AI Costs with Sparse Attention

Chinese AI startup DeepSeek releases its V3.2-Exp model, introducing a sparse attention feature that halves operational costs while improving long document handling. Despite the efficiency gains, experts still question the overall reliability and safety of the new architecture.

Hugging Face Introduces Affordable Open Source Humanoid Robots

Hugging Face unveils two new open source humanoid robots, HopeJR and Reachy Mini, aiming to make robotics accessible and affordable. The full-size HopeJR costs around $3,000 while the desktop Reachy Mini is priced between $250 and $300.

Hugging Face Launches Free Open Source Computer Agent to Rival Operator

Hugging Face introduces Open Computer Agent, a free cloud-hosted AI tool that controls a virtual Linux machine to complete basic tasks. The project aims to prove that open models can power complex agentic workflows despite current speed and accuracy limitations.

Hugging Face Acquires Pollen Robotics to Push Open-Source Humanoid Robots

Hugging Face purchases French startup Pollen Robotics, makers of the Reachy 2 humanoid robot, to expand its open-source robotics initiatives. The deal allows developers to buy the robot and directly contribute to its underlying code.

Hugging Face Acquires Pollen Robotics to Push Open-Source AI Into the Physical World

AI powerhouse Hugging Face acquires French humanoid robot maker Pollen Robotics to advance open-source embodied AI. The deal brings Pollen's Reachy 2 robot and its 20 employees under Hugging Face's umbrella.

Stability AI Launches Model That Transforms 2D Photos Into 3D Video Scenes

Stability AI introduces Stable Virtual Camera, a new generative AI tool that converts static images into immersive videos with realistic depth. The research preview is currently available for noncommercial use on Hugging Face.

Alibaba Launches Open Source Reasoning Model to Rival OpenAI's O1

Alibaba introduces QwQ-32B-Preview, an open-source reasoning AI that outperforms OpenAI's o1 on key math benchmarks. The 32.5-billion-parameter model self-fact-checks but exhibits Chinese political censorship.

Protesters Leak Access to OpenAI's Unreleased Sora Video Generator

A group of artists leaks access to OpenAI's unreleased Sora video generator to protest unpaid labor and PR manipulation. The public front end allows anyone to create 10-second videos before OpenAI shuts it down.