ggml/llama.cpp releases: b10630

TL;DR

The ggml/llama.cpp project updated its build system integration to include fixes for fp16 guard and a CMake caching issue, as part of pull request #26077. These changes enhance the project's stability and performance, particularly in handling floating-point operations.

Detailed Summary

The ggml/llama.cpp project released updates including reworking the KleidiAI build system/integration, adding an fp16 guard, and fixing a CMake caching issue. These changes were signed off by Jonathan Clohessy from ARM. The updates aim to improve the project's development workflow and stability.

Key Points

• • Rework KleidiAI Build System/Integration
• • Add fp16 guard, and fix cmake caching issue
• • Fix fo