Skip to content
Discussion options

You must be logged in to vote

The exact replacement for --no-mmap is:

llama-server ... --load-mode none

Or via the environment:

LLAMA_ARG_LOAD_MODE=none

The current argument parser maps --no-mmap directly to LLAMA_LOAD_MODE_NONE. The old flag still works for now, but is deprecated.

--load-mode mlock is different: it keeps the model locked in RAM. Also, avoid passing both the old and new flags because only the last one takes effect.

Source: current argument parser.

Replies: 2 comments

Comment options

You must be logged in to vote
0 replies
Answer selected by yves5141
Comment options

You must be logged in to vote
0 replies
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Category
Q&A
Labels
None yet
2 participants