fix cmake build (#5505)

llm/ext_server/CMakeLists.txt

@@ -3,11 +3,14 @@ set(TARGET ollama_llama_server)
 option(LLAMA_SERVER_VERBOSE "Build verbose logging option for Server" ON)
 include_directories(${CMAKE_CURRENT_SOURCE_DIR})
 add_executable(${TARGET} server.cpp utils.hpp json.hpp httplib.h)
-install(TARGETS ${TARGET} RUNTIME)
 target_compile_definitions(${TARGET} PRIVATE
     SERVER_VERBOSE=$<BOOL:${LLAMA_SERVER_VERBOSE}>
 )
 target_link_libraries(${TARGET} PRIVATE ggml llama common llava ${CMAKE_THREAD_LIBS_INIT})
+install(TARGETS ollama_llama_server ggml llama
+        RUNTIME DESTINATION "${CMAKE_BINARY_DIR}/bin"
+        LIBRARY DESTINATION "${CMAKE_BINARY_DIR}/bin"
+        COMPONENT ollama_llama_server)
 if (WIN32)
     TARGET_LINK_LIBRARIES(${TARGET} PRIVATE ws2_32)
 endif()

llm/generate/gen_common.sh

@@ -81,6 +81,7 @@ apply_patches() {
 build() {
     cmake -S ${LLAMACPP_DIR} -B ${BUILD_DIR} ${CMAKE_DEFS}
     cmake --build ${BUILD_DIR} ${CMAKE_TARGETS} -j8
+    cmake --install ${BUILD_DIR} --component ollama_llama_server
 }
 
 compress() {