How to compile torch with CUDA from git ?
from Getting started with Torch …
git clone GitHub - torch/distro: Torch installation in a self-contained folder ~/torch --recursive
cd ~/torch; bash install-deps;
./install.sh
not work :(
Help please!
-- Found Torch7 in /home/ufo/src/torch/install
-- Found CUDA: /usr/local/cuda (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): 5.3
-- got cuda version 10.0
-- Found CUDA with FP16 support, compiling with torch.CudaHalfTensor
-- CUDA_NVCC_FLAGS: -gencode;arch=compute_53,code=sm_53;-DCUDA_HAS_FP16=1
-- THC_SO_VERSION: 0
-- Performing Test HAS_LUAL_SETFUNCS
-- Performing Test HAS_LUAL_SETFUNCS - Failed
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 /home/ufo/src/torch/extra/cutorch/lib/THC
-- Configuring incomplete, errors occurred!
See also "/home/ufo/src/torch/extra/cutorch/build/CMakeFiles/CMakeOutput.log".
See also "/home/ufo/src/torch/extra/cutorch/build/CMakeFiles/CMakeError.log".
Error: Build error: Failed building.