python3 setup.py bdist_wheel error

[Nirvana93]

Hello,thanks for your code,I have following problem when running python3 setup.py bdist_wheel error:

[hpc-7-005 pytorch binding]$ python3 setup.py bdist_wheel
running bdist_wheel
runnina build
running build deps
The C compiler identification is GNU 5.4.0
The CXX compiler identification is GNU 5.4.0
Check for working c compiler: /opt/compiler/acc-5.4-os7/bin/gcc
Check for working c compiler: /opt/compiler/gcc-5.4-os7/bin/gcc --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: /opt/compiler/qcc-5.4-os7/bin/g++
Check for working CXX compiler: /opt/compiler/acc-5.4-0s7/bin/a++ -- works
Detecting CXX compiler ABI into
Detecting CXX compiler ABI info - done
Detectina CXX compile features
Detecting CXX compile features -done
Found CUDA: /opt/lib/cuda-9.1.1 (found suitable version "9.1", minimum required is "6.5"1- cuda found TRUE
CMake Warning at CMakeLists.txt:52 (FIND PACKAGE):
By not providing "FindTorch.cmake" in CMAKE MODULE PATH this project hasasked CMake to find a package configuration file provided by "Torch", butCMake did not find one.

Could not find a package configuration file provided by "Torch" with any ofthe following names:

TorchConfig.cmake
torch-config.cmake

Add the installation prefix of "Torch" to CMAKE PREFIX PATH or set"Torch DIR" to a directory containing one of the above files. If "Torch"provides a separate development package or SDK, be sure it has beeninstalled.

Torch found Torch DIR-NOTFOUND
Building shared Library with GPU support
Configuring done
Generatina done
Build files have been written to: /ps2/cv6/shuaishao3/0CR/warp-ctc-travis-ci/pytorch binding/build/lib

[16%] Building NVCC (Device) obiect CMakeFiles/warpctc.dir/src/warpctc generated_reduce.cu.o
[33%] Building NVCC (Device) obiect CMakeFiles/warpctc.dir/src/warpctc generated_ctc_entrypoint.cu.
Scanning dependencies of target warpctc
Linking CXX shared Library Libwarpctc.so
[33%] Built target warpctc
Scanning dependencies of target test cpu
[50%] Buildina CXX object CMakeFiles/test cpu.dir/tests/test cpu.cpp.o
[66%] Building CXX object CMakeFiles/test cpu.dir/tests/random.cpp.o
Linking cxx executabLe test cpu
[66%] Built target test cpu
[83%] Building NVCC (Device) obiect CMakeFiles/test gpu.dir/tests/test gpu generated test gpu.cu.ocanning dependencies of target test gpu
[100%] Building Cxx obiect CMakeFiles/test qpu.dir/tests/random.cpp.o
Linking CxX executabLe test gpu
[100%] Built target test qpu
running build py
reatina build/lib.linux-x86 64-3.6
reating build/lib.linux-x86 64-3.6/warpcto
copying warpctc/ init .py -> build/lib. Linux-x86 64-3.6/warpctc
reating build/lib.Linux-x86 64-3.6/warpctc/lib
copying warpctc/lib/libwarpctc.so -> build/lib.Linux-x86 64-3.6/warpctc/lib
runnina build ext
building 'warpctc. warpctc' extension
reating build/temp.Linux-x86 64-3.6
creating build/temp.linux-x86 64-3.6/sro
gcc -Wsign-compare-DNDEBUG -g-fwrapv -03 -WalL-Wstrict-prototypes -fPIC-I../include -I/home/cv6/huaishao3/anaconda3/lib/python3.6/site-packages/torch/lib/include -I/home/cv6/shuaishao3/anaconda3/lib/python3.6/site-packages/torch/lib/include/TH -I/home/cv6/shuaishao3/anaconda3/lib/python3.6/site-packages/torch/lib/incLude/THC -I/home/cv6/shuaishao3/anaconda3/include/python3.6m -c src/ warpctcpp -o build/temp.linux-x86 64-3.6/src/ warpctc.o-DTORCH EXTENSION NAME- warpctc-D GLIBCXX USE CXX11 ABI-O-std=C++11
cclplus: warning: command line option '-Wstrict-prototypes' is valid for C/ObiC but not for C++
src/ _warpctc.cpp:33:79: error: macro "AT_ASSERT" passed 2 arguments, but takes just 1
AT ASSERT(! input_Lengths.type().is_cuda(), "input_Lengths must be on CPU");
src/ _warpctc.cpp:39:65: error: macro "AT_ASSERT" passed 2 arguments, but takes just 1
AT ASSERT(! input_Lengths.type().is_cuda(), "input_Lengths must be on CPU");
src/ _warpctc.cpp:45:79: error: macro "AT_ASSERT" passed 2 arguments, but takes just 1
ASSERT(! Label_Lengths .type().is_cuda(), "label_Lengths must be on CPU");
src/ warpctc.cpp: In function 'std::tuple<at::Tensor, at::Tensor> ctc(at::Tensor, at::Tensor, at::Tensor, at::Tensor, int, bool)':
src/ warpctc.cpp:33:5: error: 'AT ASSERT' was not declared in this scope
AT ASSERT(! input Lengths .type().is cuda(), "input Lengths must be on CPU");
error: command 'gcc' failed with exit status 1