Models & Labs
Llama.cpp b10313 Release Adds LRU Scheduler
llama.cpp Releaseshigh confidence
Why it matters
- →The LRU scheduler improves task management efficiency in llama.cpp.
- →Expanded platform support enhances compatibility and performance.
- →The update focuses on technical improvements rather than new model architectures.
Llama.cpp has released its b10313 update, featuring the addition of an LRU scheduler to improve task management efficiency. The update addresses coalescing in the waiting queue and includes fixes for stream cases. It also expands platform support, with builds for Vulkan and ROCm 7.2 on Ubuntu, and CUDA 12 and 13 on Windows. This release focuses on enhancing performance and compatibility, though it does not introduce new model architectures.
Read originalMore from llama.cpp Releases
Open Sourcemodels
Llama.cpp b10310 Release Enhances Aarch64 Support
The latest b10310 release of llama.cpp introduces significant improvements for aarch64 architecture, particularly with the addition of HWCAP fallbacks and refined fp16 variant detection. This update ensures better compatibility and performance for devices using aarch64, such as those running on macOS Apple Silicon and various Linux distributions. By requiring HWCAP_ASIMDHP for aarch64 fp16 CPU variants, the release enhances the handling of half-precision arithmetic. While no new models are introduced, these technical adjustments make llama.cpp more robust for developers working across diverse hardware configurations.
llama.cpp Releases
Models & Labsmodels
llama.cpp b10311 Release Enhances TTS Generation
The b10311 release of llama.cpp tackles inefficiencies in text-to-speech (TTS) generation by refining how text streams are processed. Previously, the system would redundantly handle utterances, causing them to be read twice before completion. This update aligns the streaming overlay with the non-streaming prefill, effectively eliminating the duplication. Developers working with TTS systems will find this change streamlines the generation process and boosts efficiency. The update is accessible on macOS, Linux, and Windows, ensuring that a broad range of users can benefit from these improvements.
llama.cpp Releases
Open Sourcemodels
llama.cpp b10312 Release Expands Platform Support
The latest b10312 release of llama.cpp continues its trend of broadening platform compatibility, now including support for a variety of systems such as Ubuntu with ROCm 7.2 and Windows with CUDA 13.3. This update ensures that developers working across different hardware configurations, from Apple Silicon to Windows x64, have access to optimized builds. While there are no groundbreaking new features, the release solidifies llama.cpp's position as a versatile tool for AI inference across multiple systems. This means developers can now more easily integrate llama.cpp into their workflows, regardless of their preferred platform.
llama.cpp Releases
More in Models & Labs
© The Verge AIModels & Labsmodels
OpenAI Halts Astra Model Over Security Concerns
OpenAI has paused the development of its Astra model due to concerns about its cybersecurity capabilities. Internal evaluations revealed that Astra could autonomously identify and exploit zero-day vulnerabilities, raising alarms under OpenAI's Preparedness Framework. This decision reflects OpenAI's commitment to prioritizing safety and security in AI development. Although Astra was not involved in the recent Hugging Face breach, the incident has prompted OpenAI to enhance its security protocols. The company is implementing stricter controls and universal monitoring for high-capability models to prevent potential risks. This move signals a cautious approach to managing powerful AI technologies.
The Verge AI
© GitHub ChangelogModels & Labsagents
GitHub API Enhances Agent App Activity Tracking
GitHub's Copilot usage metrics API now provides detailed insights into agent app activity, allowing teams to track usage by individual agents like Claude and Codex. This update enables organizations to distinguish between different agents' activities, offering clarity on which agents are being utilized and how they compare in adoption rates. By breaking down activity per agent, teams can make informed decisions about agent rollouts and licensing based on actual usage data. This enhancement is particularly useful for enterprise owners and billing managers who need precise metrics for strategic planning.
GitHub Changelog
Models & Labsmodels
OpenAI Shares Cybersecurity Evaluations for Astra
OpenAI is actively working to enhance the cybersecurity of its Astra platform by releasing preliminary evaluations. This initiative demonstrates OpenAI's dedication to improving security measures and addressing vulnerabilities before they can be exploited. By making these evaluations available, OpenAI aims to build transparency and trust in its cybersecurity practices. This effort is part of a larger strategy to ensure Astra remains secure as it develops, setting an example for how AI platforms can responsibly manage cyber threats.
OpenAI