root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-24 10:24:35 +00:00
Code Issues Actions 5 Packages Projects Releases Wiki Activity

devops : add intel oneapi dockerfile (#5068)

Browse Source 
Co-authored-by: Xuan Son Nguyen <xuanson.nguyen@snowpack.eu>
This commit is contained in:
Xuan Son Nguyen 2024-01-23 08:11:39 +01:00 committed by GitHub
parent 125d03a503
commit 2bed4aa3f3
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 32 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
.devops/main-intel.Dockerfile Normal file
View File

@ -0,0 +1,26 @@
ARG ONEAPI_VERSION=2024.0.1-devel-ubuntu22.04
ARG UBUNTU_VERSION=22.04
FROM intel/hpckit:$ONEAPI_VERSION as build
RUN apt-get update && \
apt-get install -y git
WORKDIR /app
COPY . .
# for some reasons, "-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=Intel10_64lp -DLLAMA_NATIVE=ON" give worse performance
RUN mkdir build && \
cd build && \
cmake .. -DCMAKE_C_COMPILER=icx -DCMAKE_CXX_COMPILER=icpx && \
cmake --build . --config Release --target main server
FROM ubuntu:$UBUNTU_VERSION as runtime
COPY --from=build /app/build/bin/main /main
COPY --from=build /app/build/bin/server /server
ENV LC_ALL=C.utf8
ENTRYPOINT [ "/main" ]

1
.github/workflows/docker.yml vendored
View File

@ -35,6 +35,7 @@ jobs:
- { tag: "full-cuda", dockerfile: ".devops/full-cuda.Dockerfile", platforms: "linux/amd64" }
- { tag: "light-rocm", dockerfile: ".devops/main-rocm.Dockerfile", platforms: "linux/amd64,linux/arm64" }
- { tag: "full-rocm", dockerfile: ".devops/full-rocm.Dockerfile", platforms: "linux/amd64,linux/arm64" }
- { tag: "light-intel", dockerfile: ".devops/main-intel.Dockerfile", platforms: "linux/amd64" }
steps:
- name: Check out the repo
uses: actions/checkout@v3

5
CMakeLists.txt
View File

@ -478,6 +478,11 @@ function(get_flags CCID CCVER)
if (CCVER VERSION_GREATER_EQUAL 8.1.0)
set(CXX_FLAGS ${CXX_FLAGS} -Wextra-semi)
endif()
elseif (CCID MATCHES "Intel")
# enable max optimization level when using Intel compiler
set(C_FLAGS -ipo -O3 -static -fp-model=fast -flto -fno-stack-protector)
set(CXX_FLAGS -ipo -O3 -static -fp-model=fast -flto -fno-stack-protector)
add_link_options(-fuse-ld=lld -static-intel)
endif()
set(GF_C_FLAGS ${C_FLAGS} PARENT_SCOPE)