Products News

Hugging Face

Waypoint-1.5: Higher-Fidelity Interactive Worlds for Everyday GPUs

Products

Waypoint-1.5 enables high-quality interactive 3D environments to run efficiently on consumer-grade GPUs, making advanced simulation and AI training tools accessible beyond specialized data centers. This democratization of world-building technology matters because it lowers barriers for researchers, developers, and companies building embodied AI systems, robotic simulations, and game engines—work previously constrained by expensive hardware requirements.

Read more
Hugging Face

Multimodal Embedding & Reranker Models with Sentence Transformers

Products

Sentence Transformers has introduced new multimodal embedding and reranking capabilities, expanding the library's functionality beyond text to handle images, audio, and other data types simultaneously. This development matters because it democratizes access to powerful cross-modal AI tools that enterprises typically build in-house, enabling developers to more easily build search systems, recommendation engines, and retrieval-augmented generation (RAG) pipelines that work across different data types without expensive custom engineering.

Read more
Hugging Face

Safetensors is Joining the PyTorch Foundation

Products

Safetensors, the lightweight file format developed by Hugging Face for safely storing and sharing machine learning model weights, is now joining the PyTorch Foundation as an official project. This move legitimizes Safetensors as an industry standard and signals the AI community's commitment to making model distribution more secure and efficient across frameworks beyond just PyTorch. For practitioners and organizations, this means broader adoption, better maintenance, and increased confidence that Safetensors will remain the go-to format for handling the massive model files that power modern AI applications.

Read more
Ars Technica

Thousands of consumer routers hacked by Russia's military

Products

Russian military hackers have compromised thousands of consumer routers worldwide, potentially enabling large-scale surveillance, data theft, and network infiltration for espionage purposes. This incident matters to AI followers because compromised home and small business networks create backdoors that could be exploited to access AI training data, proprietary models, and cloud services that increasingly power both consumer and enterprise AI systems.

Read more
PreviousPage 2 of 5Next