Installing cutorch with CUDA 10.0

doneforaiur commented 5 years ago

$luarocks install cutorchgives:

Installing https://raw.githubusercontent.com/torch/rocks/master/cutorch-scm-1.rockspec...
Using https://raw.githubusercontent.com/torch/rocks/master/cutorch-scm-1.rockspec... switching to 'build' mode
'cutorch' dizinine çoğaltılıyor...
remote: Enumerating objects: 229, done.
remote: Counting objects: 100% (229/229), done.
remote: Compressing objects: 100% (184/184), done.
remote: Total 229 (delta 62), reused 90 (delta 43), pack-reused 0
Nesneler alınıyor: 100% (229/229), 241.83 KiB | 248.00 KiB/s, bitti.
Farklar çözülüyor: 100% (62/62), bitti.
Warning: unmatched variable LUALIB

jopts=$(getconf _NPROCESSORS_CONF)

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

Building on 4 cores
-- The C compiler identification is GNU 7.3.0
-- The CXX compiler identification is GNU 7.3.0
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Found Torch7 in /home/lelu/torch/install
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Looking for pthread_create
-- Looking for pthread_create - not found
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - found
-- Found Threads: TRUE  
-- Found CUDA: /usr/local/cuda-10.0 (found suitable version "10.0", minimum required is "6.5") 
-- Removing -DNDEBUG from compile flags
-- TH_LIBRARIES: TH
-- MAGMA not found. Compiling without MAGMA support
-- Autodetected CUDA architecture(s): 5.0 
-- got cuda version 10.0
-- Found CUDA with FP16 support, compiling with torch.CudaHalfTensor
-- CUDA_NVCC_FLAGS: -D__CUDA_NO_HALF_OPERATORS__;-gencode;arch=compute_50,code=sm_50;-DCUDA_HAS_FP16=1
-- THC_SO_VERSION: 0
-- Performing Test HAS_LUAL_SETFUNCS
-- Performing Test HAS_LUAL_SETFUNCS - Success
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_device_LIBRARY (ADVANCED)
    linked by target "THC" in directory /tmp/luarocks_cutorch-scm-1-399/cutorch/lib/THC

-- Configuring incomplete, errors occurred!
See also "/tmp/luarocks_cutorch-scm-1-399/cutorch/build/CMakeFiles/CMakeOutput.log".
See also "/tmp/luarocks_cutorch-scm-1-399/cutorch/build/CMakeFiles/CMakeError.log".

Error: Build error: Failed building.

$nvidia-smigives;

Sun Nov 18 14:23:24 2018       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 410.72       Driver Version: 410.72       CUDA Version: 10.0     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 950M    Off  | 00000000:04:00.0 Off |                  N/A |
| N/A   48C    P8    N/A /  N/A |      0MiB /  4046MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+

Updating torch didn't solve the problem.

ghost commented 5 years ago

Exactly the same issue with you. I tried

remove old CMake, and get CMake 13.0 from their websit, but still didn't solve the problem
remove CUDA 10.0, go back to CUDA 9.0, it works!! (remember to remove all packages, update sourcelist, repair PATH, etc)

And, here is a new solution for cutorch+CUDA 10 . Hope it works! Last and last, after all set up work, remember to set CUDA_VISIBLE_DEVICES varialbe in order to select a device

mrgloom commented 4 years ago

sudo luarocks install cutorch

Error: No results matching query were found.

torch / cutorch

Installing cutorch with CUDA 10.0 #836