The ggml-org team has released llama.cpp b11598, with a change to load YaRN beta metadata for Laguna models. The release page lists downloads for macOS, iOS and Linux, including CPU and Vulkan builds for Linux.
The release notes do not explain what user-visible difference the metadata change makes. They also do not list benchmark results or further details about Laguna model support.
Why it matters
People running Laguna models with llama.cpp have a release that loads their YaRN beta metadata, though the notes do not say what practical change they should expect.
