Because we specify a minimum CMake version of 3.21 in the llmodel project (gpt4all-backend), we get the NEW behavior of CMP0104, which causes enable_langauge(CUDA) to set CMAKE_CUDA_ARCHITECTURES t...
enable_language(CUDA) if(NOT USER_DEFINED_CMAKE_CUDA_ARCHITECTURES) set(CMAKE_CUDA_ARCHITECTURES "") endif() elseif(UNIX) message(WARNING "CUDA: Not detected! If you are not using the default host compiler (g++) then you need to specify both CMAKE_CUDA_HOST_COMPILER and CMAKE_CUDA_COMP...
- # Seehttps://cmake.org/cmake/help/latest/policy/CMP0104.htmlfor the - # default CUDA architecture for each version of CUDA. - set(CMAKE_CUDA_ARCHITECTURES OFF) + if(NOT DEFINED CMAKE_CUDA_ARCHITECTURES) + message(STATUS "CMAKE_CUDA_ARCHITECTURES not set, using native") + set(CMAKE...