ollama/llm/ext_server/CMakeLists.txt


set(TARGET ext_server)
option(LLAMA_SERVER_VERBOSE "Build verbose logging option for Server" ON)
if (WIN32)
    add_library(${TARGET} SHARED ext_server.cpp ../llama.cpp/llama.cpp)
else()
    add_library(${TARGET} STATIC ext_server.cpp ../llama.cpp/llama.cpp)
endif()
target_compile_features(${TARGET} PRIVATE cxx_std_11)
target_compile_definitions(${TARGET} PUBLIC LLAMA_SERVER_LIBRARY=1)
target_link_libraries(${TARGET} PRIVATE ggml llava common )
set_target_properties(${TARGET} PROPERTIES POSITION_INDEPENDENT_CODE ON)
target_compile_definitions(${TARGET} PRIVATE SERVER_VERBOSE=$<BOOL:${LLAMA_SERVER_VERBOSE}>)
install(TARGETS ext_server LIBRARY)

if (CUDAToolkit_FOUND)
    target_include_directories(${TARGET} PRIVATE ${CMAKE_CUDA_TOOLKIT_INCLUDE_DIRECTORIES})
    if (WIN32)
        target_link_libraries(${TARGET} PRIVATE nvml)
    endif()
endif()
Refactor how we augment llama.cpp This changes the model for llama.cpp inclusion so we're not applying a patch, but instead have the C++ code directly in the ollama tree, which should make it easier to refine and update over time. 2023-12-22 17:51:53 +00:00
			`set(TARGET ext_server)`
			`option(LLAMA_SERVER_VERBOSE "Build verbose logging option for Server" ON)`
Add multiple CPU variants for Intel Mac This also refines the build process for the ext_server build. 2024-01-13 00:28:00 +00:00			`if (WIN32)`
Adapt our build for imported server.cpp 2024-03-12 20:51:44 +00:00			`add_library(${TARGET} SHARED ext_server.cpp ../llama.cpp/llama.cpp)`
Add multiple CPU variants for Intel Mac This also refines the build process for the ext_server build. 2024-01-13 00:28:00 +00:00			`else()`
Adapt our build for imported server.cpp 2024-03-12 20:51:44 +00:00			`add_library(${TARGET} STATIC ext_server.cpp ../llama.cpp/llama.cpp)`
Add multiple CPU variants for Intel Mac This also refines the build process for the ext_server build. 2024-01-13 00:28:00 +00:00			`endif()`
Refactor how we augment llama.cpp This changes the model for llama.cpp inclusion so we're not applying a patch, but instead have the C++ code directly in the ollama tree, which should make it easier to refine and update over time. 2023-12-22 17:51:53 +00:00			`target_compile_features(${TARGET} PRIVATE cxx_std_11)`
			`target_compile_definitions(${TARGET} PUBLIC LLAMA_SERVER_LIBRARY=1)`
Add multiple CPU variants for Intel Mac This also refines the build process for the ext_server build. 2024-01-13 00:28:00 +00:00			`target_link_libraries(${TARGET} PRIVATE ggml llava common )`
			`set_target_properties(${TARGET} PROPERTIES POSITION_INDEPENDENT_CODE ON)`
			`target_compile_definitions(${TARGET} PRIVATE SERVER_VERBOSE=$<BOOL:${LLAMA_SERVER_VERBOSE}>)`
			`install(TARGETS ext_server LIBRARY)`
Refactor how we augment llama.cpp This changes the model for llama.cpp inclusion so we're not applying a patch, but instead have the C++ code directly in the ollama tree, which should make it easier to refine and update over time. 2023-12-22 17:51:53 +00:00
			`if (CUDAToolkit_FOUND)`
			`target_include_directories(${TARGET} PRIVATE ${CMAKE_CUDA_TOOLKIT_INCLUDE_DIRECTORIES})`
			`if (WIN32)`
Add multiple CPU variants for Intel Mac This also refines the build process for the ext_server build. 2024-01-13 00:28:00 +00:00			`target_link_libraries(${TARGET} PRIVATE nvml)`
Refactor how we augment llama.cpp This changes the model for llama.cpp inclusion so we're not applying a patch, but instead have the C++ code directly in the ollama tree, which should make it easier to refine and update over time. 2023-12-22 17:51:53 +00:00			`endif()`
			`endif()`