Llama.cpp安装

参考：https://github.com/ggml-org/llama.cpp/blob/master/docs/build.md

1、下载源码

git clone https://github.com/ggml-org/llama.cpp
cd llama.cpp
apt install cmake

2、编译源码

cmake -B build -DGGML_CUDA=ON
cmake --build build --config Release

出现错误：

Could NOT find CURL.  Hint: to disable this feature, set -DLLAMA_CURL=OFF

出现下面错误
base) ywj@AI:~/software/github_project/llama.cpp$ cmake -B build -DGGML_CUDA=ON
-- The C compiler identification is GNU 11.4.0
-- The CXX compiler identification is GNU 11.4.0
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /usr/bin/cc - skipped
-- Detecting C compile features
-- Detecting C compile features - done
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/c++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Found Git: /usr/bin/git (found version "2.34.1") 
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Success
-- Found Threads: TRUE  
-- Warning: ccache not found - consider installing it for faster compilation or disable this warning with GGML_CCACHE=OFF
-- CMAKE_SYSTEM_PROCESSOR: x86_64
-- Including CPU backend
-- Found OpenMP_C: -fopenmp (found version "4.5") 
-- Found OpenMP_CXX: -fopenmp (found version "4.5") 
-- Found OpenMP: TRUE (found version "4.5")  
-- x86 detected
-- Adding CPU backend variant ggml-cpu: -march=native 
-- Found CUDAToolkit: /usr/local/cuda-12.2/include (found version "12.2.140") 
-- CUDA Toolkit found
-- Using CUDA architectures: 50-virtual;61-virtual;70-virtual;75-virtual;80-virtual;86-real;89-real
-- The CUDA compiler identification is NVIDIA 12.2.140
-- Detecting CUDA compiler ABI info
-- Detecting CUDA compiler ABI info - done
-- Check for working CUDA compiler: /usr/local/cuda-12.2/bin/nvcc - skipped
-- Detecting CUDA compile features
-- Detecting CUDA compile features - done
-- CUDA host compiler is GNU 11.4.0
-- Including CUDA backend
-- Could NOT find CURL (missing: CURL_LIBRARY CURL_INCLUDE_DIR) 
CMake Error at common/CMakeLists.txt:98 (message):
  Could NOT find CURL.  Hint: to disable this feature, set -DLLAMA_CURL=OFF


-- Configuring incomplete, errors occurred!
See also "/home/ywj/software/github_project/llama.cpp/build/CMakeFiles/CMakeOutput.log".

执行：

cmake -B build -DGGML_CUDA=ON -DLLAMA_CURL=OFF

cmake -B build -DGGML_CUDA=ON -DLLAMA_CURL=OFF
-- Warning: ccache not found - consider installing it for faster compilation or disable this warning with GGML_CCACHE=OFF
-- CMAKE_SYSTEM_PROCESSOR: x86_64
-- Including CPU backend
-- x86 detected
-- Adding CPU backend variant ggml-cpu: -march=native 
-- CUDA Toolkit found
-- Using CUDA architectures: 50-virtual;61-virtual;70-virtual;75-virtual;80-virtual;86-real;89-real
-- CUDA host compiler is GNU 11.4.0
-- Including CUDA backend
-- Configuring done
-- Generating done
-- Build files have been written to: /home/ywj/software/github_project/llama.cpp/build

cmake --build build --config Release
[  0%] Building C object ggml/src/CMakeFiles/ggml-base.dir/ggml.c.o
[  1%] Building C object ggml/src/CMakeFiles/ggml-base.dir/ggml-alloc.c.o
[  1%] Building CXX object ggml/src/CMakeFiles/ggml-base.dir/ggml-backend.cpp.o
[  1%] Building CXX object ggml/src/CMakeFiles/ggml-base.dir/ggml-opt.cpp.o
[  2%] Building CXX object ggml/src/CMakeFiles/ggml-base.dir/ggml-threading.cpp.o
[  2%] Building C object ggml/src/CMakeFiles/ggml-base.dir/ggml-quants.c.o
[  2%] Building CXX object ggml/src/CMakeFiles/ggml-base.dir/gguf.cpp.o
[  3%] Linking CXX shared library ../../bin/libggml-base.so
[  3%] Built target ggml-base
[  3%] Building CUDA object ggml/src/ggml-cuda/CMakeFiles/ggml-cuda.dir/acc.cu.o
[  3%] Building CUDA object ggml/src/ggml-cuda/CMakeFiles/ggml-cuda.dir/arange.cu.o
[  4%] Building CUDA object ggml/src/ggml-cuda/CMakeFiles/ggml-cuda.dir/argmax.cu.o
[  4%] Building CUDA object ggml/src/ggml-cuda/CMakeFiles/ggml-cuda.dir/argsort.cu.o
[  4%] Building CUDA object ggml/src/ggml-cuda/CMakeFiles/ggml-cuda.dir/binbcast.cu.o
[  5%] Building CUDA object ggml/src/ggml-cuda/CMakeFiles/ggml-cuda.dir/clamp.cu.o
[  5%] Building CUDA object ggml/src/ggml-cuda/CMakeFiles/ggml-cuda.dir/concat.cu.o
[  5%] Building CUDA object ggml/src/ggml-cuda/CMakeFiles/ggml-cuda.dir/conv-transpose-1d.cu.o

执行完 cmake --build build --config Release ，就等着吧，这个编译过程需要点时间