llama.cpp b11385: Fixes Inactive Preset Argument Key in Server Component
A small server-side fix removes a non-functional configuration key from the allowed parameter list in llama.cpp, clarifying configuration and reducing confusion for developers deploying custom models.
What changed?
The b11385 release of llama.cpp fixes an issue in the server component by removing a dead configuration key (LLAMA_ARG_HF_REPO_FILE) from the preset allow-list. This key was previously accepted by the configuration but had no effect or support in the actual server implementation.
Why does it matter to an everyday developer?
If you've been trying to use the LLAMA_ARG_HF_REPO_FILE argument in your llama.cpp server deployment or scripts, this release clarifies that the parameter is not supported and has no functional impact. This reduces configuration confusion and helps ensure that only active, supported settings are used in model deployments.
What can the developer do now?
Developers should check their llama.cpp server configuration files and deployment scripts to ensure they are not using LLAMA_ARG_HF_REPO_FILE, as it is now clearly unsupported and ignored. No code changes are required unless you were relying on this option. Pre-built binaries for all major operating systems and architectures are available for immediate update.