llama.cpp b9596 release cleans up server router logs

The ggml-org team has released llama.cpp b9596, the latest version of their widely-used open-source C/C++ inference engine for large language models. This update focuses on improving the server component, specifically when operating in router mode. The commit message 'server: skip unused log lines on router mode (#24463)' indicates that the server now filters out log lines that are irrelevant when routing requests to different models, making logs cleaner and more useful for debugging and monitoring.

The release comes with an extensive set of pre-built binaries covering nearly every major platform. macOS users get builds for Apple Silicon (arm64), Intel (x64), and iOS as an XCFramework. Linux supports Ubuntu x64, arm64, and s390x, with acceleration backends including Vulkan, ROCm 7.2, OpenVINO, and SYCL (FP32). Android gets an arm64 CPU build. Windows offers x64 and arm64 CPU builds, plus GPU-accelerated versions with CUDA 12 and 13, Vulkan, SYCL, and HIP. openEuler systems (both x86 and aarch64) are also supported with ACL Graph optimizations. This breadth ensures developers and system administrators can deploy llama.cpp on virtually any infrastructure, from cloud servers to edge devices.