The ggml-org team has published llama.cpp b11848, a new release of the open-source software. Its listed change allows concurrent downloads across processes.
The release page offers builds for macOS, iOS and Linux, including CPU and Vulkan options for some Linux systems. It does not explain which downloads are affected or quantify the change.
Why it matters
For people running multiple llama.cpp processes, the change may allow downloads to overlap; the release notes do not specify the affected workflow or its practical effect.
