]> git.djapps.eu Git - pkg/ggml/sources/llama.cpp/commit
overview / pkg / ggml / sources / llama.cpp / commit
summary | shortlog | log | commit | commitdiff | tree
(parent: 27c8bb4)
common : allow --offline in llama download (#25091)
authorAdrien Gallouët <redacted>
Sun, 28 Jun 2026 10:34:11 +0000 (12:34 +0200)
committerGitHub <redacted>
Sun, 28 Jun 2026 10:34:11 +0000 (12:34 +0200)
commitc1a1c8ee94e37ff7ba2c872e783aaf7f77e0f320
treec561ca2ba886f28a3b03274a4b837ff642ff776dtree
parent27c8bb4f63ad9f20bf5901067810a4be5ffe20c4commit | diff
common : allow --offline in llama download (#25091)

Expose the existing --offline flag to `llama download` so a script can
run it to check whether a model is already cached and ready to be served
without touching the network.

Also fix a latent use-after-free in the URL-task on_done callback:
first_path is block-scoped and was captured by reference, but invoked
after the block ends.

Signed-off-by: Adrien Gallouët <redacted>
common/arg.cpp diff | blob | history
Packaging of ggml-org/llama.cpp