Models & Labs
llama.cpp b9519 release enhances SYCL support
llama.cpp Releaseshigh confidence
Why it matters
- →Enhances performance by optimizing weight reading in SYCL backend.
- →Expands compatibility across different hardware platforms.
- →Continues to refine llama.cpp's capabilities for developers.
The b9519 release of llama.cpp introduces enhancements to its SYCL backend by porting multi-column MMVQ optimizations from the CUDA backend. This update optimizes weight reading, reducing it from once per column to once per dispatch, which is expected to improve performance for standard quantization types. While some IQ types are excluded due to compatibility issues, the release broadens llama.cpp's applicability across various hardware configurations. This update underscores llama.cpp's commitment to improving performance and compatibility across diverse computing environments.
Read originalMore from llama.cpp Releases
Open Sourcecoding
llama.cpp b9503 release focuses on Gemma 4 fix
The b9503 release of llama.cpp addresses a technical issue with the Gemma 4 audio projector embedding size, enhancing its functionality. By removing the projection_dim from clip_n_mmproj_embd, the update streamlines the codebase. This release ensures better compatibility across macOS, Linux, and Windows, with specific builds for Apple Silicon, ROCm 7.2, and CUDA 12 and 13. While it doesn't introduce new features, the update reflects a commitment to improving the software's reliability and performance. This release is a technical refinement, focusing on stability rather than groundbreaking changes.
llama.cpp Releases
Open Sourcemodels
llama.cpp b9504 release expands platform support
The b9504 release of llama.cpp continues to broaden its reach, enhancing compatibility across multiple environments. This update notably includes support for Ubuntu with ROCm 7.2, which boosts performance for AMD GPU users. While features like KleidiAI on macOS and SYCL on Windows are not yet active, the release still represents a significant step in making llama.cpp a more adaptable tool for developers. By focusing on expanding compatibility and improving the runtime experience, this update strengthens llama.cpp's position as a versatile option for developers working with different systems.
llama.cpp Releases
Open Sourcemodels
llama.cpp b9505 release expands platform support
The b9505 release of llama.cpp continues its trend of broadening compatibility across various systems, though with some notable exceptions. While macOS Apple Silicon users see KleidiAI support disabled, the release strengthens its presence on Windows with CUDA 12 and 13 DLLs, and extends Vulkan support to more environments. The inclusion of ROCm 7.2 for Ubuntu x64 users further narrows the gap between AMD and NVIDIA GPU support. This update underscores llama.cpp's commitment to being a versatile inference runtime, though some features remain disabled, indicating ongoing development challenges.
llama.cpp Releases
More in Models & Labs
Models & Labsmodels
v0.22.1 fixes CUTLASS fmin compatibility
The v0.22.1 release of vLLM addresses a critical compatibility issue with CUTLASS fmin during the initialization of DeepSeek-V4. This update ensures that users relying on this configuration experience smoother integration and improved functionality. By resolving this specific technical challenge, the release contributes to the ongoing refinement and stability of the vLLM framework. Users can now expect enhanced performance and fewer compatibility problems, reinforcing the platform's reliability. This update is a testament to the continuous efforts to maintain and improve the technical robustness of vLLM.
vLLM Releases
© TechCrunch AIModels & Labsmodels
Airbnb's Brian Chesky to Launch New AI Lab
Airbnb CEO Brian Chesky is making a strategic move into AI by backing a new lab, indicating his shift from an advisory role to a more hands-on approach in AI development. This decision stems from his dissatisfaction with current AI models and his desire to innovate in user interaction and design, areas he has prioritized at Airbnb. Although Chesky will continue as Airbnb's CEO, the new lab will operate under different leadership, tasked with competing against established AI labs. This initiative could bring new perspectives to AI, particularly in enhancing user experiences, and potentially disrupt the current landscape.
TechCrunch AI
© GitHub ChangelogModels & Labsagents
GitHub Launches Agent Tasks REST API for Copilot
GitHub has introduced a new Agent tasks REST API for Copilot Pro, Pro+, and Max users, now available in public preview. This API allows developers to programmatically initiate and monitor Copilot cloud agent tasks, integrating seamlessly into custom automation workflows. The Copilot cloud agent operates independently, making and validating code changes before submitting pull requests. This development empowers users to automate complex tasks like refactoring across multiple repositories or setting up new ones with ease. The API supports various authentication methods, enhancing its accessibility for developers.
GitHub Changelog