-
Notifications
You must be signed in to change notification settings - Fork 9.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
cuda : rename build flag to LLAMA_CUDA #6299
Conversation
This also renames some targets in the RPM packages. cc @jboero |
Thanks for the heads-up. I can switch this easy enough. |
This is being changed in Make. ggerganov#6299
Not sure why the nix CI is not completing, it seems to have timed out after 6 hours. The same things seems to be happening on master, so I don't think this PR is the cause. |
I think starting with this PR the nix CI is rebuilding ROCm each time (and timing out after 6 hours each time). In the previous merged PR #6298 and before nix ubuntu build was taking 22-23 minutes. |
You can see the previous commit in master the nix run being cancelled after 5 hours (when this was merged). |
Following upstream here: ggerganov/llama.cpp#6299 . Signed-off-by: Eric Joldasov <bratishkaerik@landless-city.net>
LLAMA_CUBLAS
still works for now to avoid breaking build scripts, but it prints a warning announcing that it will be removed in the future.GGML_USE_CUBLAS
is also renamed toGGML_USE_CUDA
.