Skip to content

[llamacpp] replace deprecated flags, better IO flag handling - #2833

Draft
bitgamma wants to merge 1 commit into
mainfrom
bitgamma/llamcpp-args
Draft

[llamacpp] replace deprecated flags, better IO flag handling#2833
bitgamma wants to merge 1 commit into
mainfrom
bitgamma/llamcpp-args

Conversation

@bitgamma

Copy link
Copy Markdown
Member

replace some deprecated llamacpp flags with their new version. It also handles better the IO flag overrides so that the defaults do not accidentally override whatever the user sets.

It also allows the user control the mmproj offload (when using GPU) since this is not in actual conflict with what we set.

since --load-mode was introduced July 23rd, we need to bump llamacpp version before merging this.

finally, a couple of unused constants (they are defined elsewhere) have been removed

@github-actions github-actions Bot added engine::llamacpp llama.cpp backend (LlamaCppServer); GPU/CPU LLM inference (Vulkan, ROCm, Metal) enhancement New feature or request labels Jul 28, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

engine::llamacpp llama.cpp backend (LlamaCppServer); GPU/CPU LLM inference (Vulkan, ROCm, Metal) enhancement New feature or request

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant