Error occurred when I install torch with the env of nvidia-docker run tensorflow image

[whytin]

I use nvidia-docker to run tensorflow/tensorflow:latest-gpu image. And then I want to install torch. Some problems happened. Env: Ubuntu16.04、Cuda 9.0、 Cudnn 7.0.5、 GPU:1080ti How can I figure out this? Thanks!

Error flew out in process:

Found CUDA on your machine. Installing CUDA packages Warning: unmatched variable LUALIB

jopts=$(getconf _NPROCESSORS_CONF)

echo "Building on $jopts cores" cmake -E make_directory build && cd build && cmake .. -DLUALIB= -DLUA_INCDIR=/distro/install/include -DCMAKE_CXX_FLAGS=${CMAKE_CXX_FLAGS} -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH="/distro/install/bin/.." -DCMAKE_INSTALL_PREFIX="/distro/install/lib/luarocks/rocks/cutorch/scm-1" && make -j$jopts install

Building on 12 cores -- Found Torch7 in /distro/install -- Removing -DNDEBUG from compile flags -- TH_LIBRARIES: TH -- MAGMA not found. Compiling without MAGMA support -- Autodetected CUDA architecture(s): 6.1 -- got cuda version 9.0 -- Found CUDA with FP16 support, compiling with torch.CudaHalfTensor -- CUDA_NVCC_FLAGS: -gencode;arch=compute_61,code=sm_61;-DCUDA_HAS_FP16=1 -- THC_SO_VERSION: 0 CMake Error: The following variables are used in this project, but they are set to NOTFOUND. Please set them or make sure they are set and tested correctly in the CMake files: CUDA_cublas_LIBRARY (ADVANCED) linked by target "THC" in directory /distro/extra/cutorch/lib/THC CUDA_cublas_device_LIBRARY (ADVANCED) linked by target "THC" in directory /distro/extra/cutorch/lib/THC CUDA_curand_LIBRARY (ADVANCED) linked by target "THC" in directory /distro/extra/cutorch/lib/THC CUDA_cusparse_LIBRARY (ADVANCED) linked by target "THC" in directory /distro/extra/cutorch/lib/THC

-- Configuring incomplete, errors occurred! See also "/distro/extra/cutorch/build/CMakeFiles/CMakeOutput.log". See also "/distro/extra/cutorch/build/CMakeFiles/CMakeError.log".

Error: Build error: Failed building.

And Cat with CMakeError.log:

Performing C SOURCE FILE Test HAS_LUAL_SETFUNCS failed with the following output: Change Dir: /distro/extra/cutorch/build/CMakeFiles/CMakeTmp

Run Build Command:"/usr/bin/make" "cmTC_74126/fast" /usr/bin/make -f CMakeFiles/cmTC_74126.dir/build.make CMakeFiles/cmTC_74126.dir/build make[1]: Entering directory '/distro/extra/cutorch/build/CMakeFiles/CMakeTmp' Building C object CMakeFiles/cmTC_74126.dir/src.c.o /usr/bin/cc -I/distro/install/include -std=c99 -Werror=implicit-function-declaration -DHAS_LUAL_SETFUNCS -o CMakeFiles/cmTC_74126.dir/src.c.o -c /distro/extra/cutorch/build/CMakeFiles/CMakeTmp/src.c /distro/extra/cutorch/build/CMakeFiles/CMakeTmp/src.c: In function 'main': /distro/extra/cutorch/build/CMakeFiles/CMakeTmp/src.c:5:20: error: 'luaL_setfuncs' undeclared (first use in this function) long i = sizeof(&luaL_setfuncs); ^ /distro/extra/cutorch/build/CMakeFiles/CMakeTmp/src.c:5:20: note: each undeclared identifier is reported only once for each function it appears in CMakeFiles/cmTC_74126.dir/build.make:65: recipe for target 'CMakeFiles/cmTC_74126.dir/src.c.o' failed make[1]: [CMakeFiles/cmTC_74126.dir/src.c.o] Error 1 make[1]: Leaving directory '/distro/extra/cutorch/build/CMakeFiles/CMakeTmp' Makefile:126: recipe for target 'cmTC_74126/fast' failed make: [cmTC_74126/fast] Error 2

Source file was:

include

int main() { long i = sizeof(&luaL_setfuncs); return 0; }

[greed2411]

@whytinI found this nvidia-docker image to be useful, ufoym/deepo. Let me know if it works out.

[whytin]

@greed2411 It did work for me, Thank you!