NVIDIA / NVIDIA/apex

Build error: error: identifier "cublasLtMatmulDescOpaque_t" is undefined

Open
#1,091 8 comments 3 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
9k
Forks
1.5k
Avg merge
2d 4h
Merged PRs (30d)
3

Description

I am using PyTorch 1.8.1 in Cuda 10.2. whenever I try to build apex I get build error.

One of the error was "error: identifier "cublasLtMatmulDescOpaque_t" is undefined". How can I ass cublas?
Build Log:

2021-04-25T17:22:28,441 Using pip 21.1 from /home/partha9/py38_apex/lib/python3.8/site-packages/pip (python 3.8)
2021-04-25T17:22:28,444 Non-user install because user site-packages disabled
2021-04-25T17:22:28,524 Created temporary directory: /tmp/pip-ephem-wheel-cache-455uizkm
2021-04-25T17:22:28,525 Created temporary directory: /tmp/pip-req-tracker-f7d4hblw
2021-04-25T17:22:28,525 Initialized build tracking at /tmp/pip-req-tracker-f7d4hblw
2021-04-25T17:22:28,525 Created build tracker: /tmp/pip-req-tracker-f7d4hblw
2021-04-25T17:22:28,525 Entered build tracker: /tmp/pip-req-tracker-f7d4hblw
2021-04-25T17:22:28,525 Created temporary directory: /tmp/pip-install-1kyf02s9
2021-04-25T17:22:28,548 Looking in links: /cvmfs/soft.computecanada.ca/custom/python/wheelhouse/nix/avx2, /cvmfs/soft.computecanada.ca/custom/python/wheelhouse/nix/generic, /cvmfs/soft.computecanada.ca/custom/python/wheelhouse/generic
2021-04-25T17:22:28,551 Processing /home/partha9/apex 2021-04-25T17:22:28,551   Created temporary directory: /tmp/pip-req-build-8zq6h56v
2021-04-25T17:22:28,551   DEPRECATION: A future pip version will change local packages to be built in-place without first copying to a temporary directory. We recommend you use --use-feature=in-tree-build to test your packages with this new behavior before it becomes the default.
2021-04-25T17:22:28,551    pip 21.3 will remove support for this functionality. You can find discussion regarding this at https://github.com/pypa/pip/issues/7555.
2021-04-25T17:22:28,990   Added file:///home/partha9/apex to build tracker '/tmp/pip-req-tracker-f7d4hblw'
2021-04-25T17:22:28,991     Running setup.py (path:/tmp/pip-req-build-8zq6h56v/setup.py) egg_info for package from file:///home/partha9/apex
2021-04-25T17:22:28,991     Created temporary directory: /tmp/pip-pip-egg-info-8ertna46
2021-04-25T17:22:28,991     Running command python setup.py egg_info
2021-04-25T17:22:30,938     No CUDA runtime is found, using CUDA_HOME='/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89'

2021-04-25T17:22:30,939     Warning: Torch did not find available GPUs on this system.
2021-04-25T17:22:30,939      If your intention is to cross-compile, this is not an error.
2021-04-25T17:22:30,939     By default, Apex will cross-compile for Pascal (compute capabilities 6.0, 6.1, 6.2),
2021-04-25T17:22:30,939     Volta (compute capability 7.0), Turing (compute capability 7.5),
2021-04-25T17:22:30,939     and, if the CUDA version is >= 11.0, Ampere (compute capability 8.0).
2021-04-25T17:22:30,939     If you wish to cross-compile for a single specific architecture,
2021-04-25T17:22:30,939     export TORCH_CUDA_ARCH_LIST="compute capability" before running setup.py.



2021-04-25T17:22:30,939     torch.__version__  = 1.8.1+cu102


2021-04-25T17:22:30,940     running egg_info
2021-04-25T17:22:30,940     creating /tmp/pip-pip-egg-info-8ertna46/apex.egg-info
2021-04-25T17:22:30,940     writing /tmp/pip-pip-egg-info-8ertna46/apex.egg-info/PKG-INFO
2021-04-25T17:22:30,940     writing dependency_links to /tmp/pip-pip-egg-info-8ertna46/apex.egg-info/dependency_links.txt
2021-04-25T17:22:30,940     writing top-level names to /tmp/pip-pip-egg-info-8ertna46/apex.egg-info/top_level.txt
2021-04-25T17:22:30,940     writing manifest file '/tmp/pip-pip-egg-info-8ertna46/apex.egg-info/SOURCES.txt'
2021-04-25T17:22:30,940     adding license file 'LICENSE' (matched pattern 'LICEN[CS]E*')
2021-04-25T17:22:30,940     reading manifest file '/tmp/pip-pip-egg-info-8ertna46/apex.egg-info/SOURCES.txt'
2021-04-25T17:22:30,941     writing manifest file '/tmp/pip-pip-egg-info-8ertna46/apex.egg-info/SOURCES.txt'
2021-04-25T17:22:30,942     /tmp/pip-req-build-8zq6h56v/setup.py:67: UserWarning: Option --pyprof not specified. Not installing PyProf dependencies!
2021-04-25T17:22:30,942       warnings.warn("Option --pyprof not specified. Not installing PyProf dependencies!")
2021-04-25T17:22:31,072   Source in /tmp/pip-req-build-8zq6h56v has version 0.1, which satisfies requirement apex==0.1 from file:///home/partha9/apex
2021-04-25T17:22:31,073   Removed apex==0.1 from file:///home/partha9/apex from build tracker '/tmp/pip-req-tracker-f7d4hblw'
2021-04-25T17:22:31,074 Created temporary directory: /tmp/pip-unpack-sf08mc58
2021-04-25T17:22:31,074 Skipping wheel build for apex, due to binaries being disabled for it.
2021-04-25T17:22:31,573 Installing collected packages: apex
2021-04-25T17:22:31,576   Created temporary directory: /tmp/pip-record-qf8wx_y_
2021-04-25T17:22:31,576     Running command /home/partha9/py38_apex/bin/python3 -u -c 'import io, os, sys, setuptools, tokenize; sys.argv[0] = '"'"'/tmp/pip-req-build-8zq6h56v/setup.py'"'"'; __file__='"'"'/tmp/pip-req-build-8zq6h56v/setup.py'"'"';f = getattr(tokenize, '"'"'open'"'"', open)(__file__) if os.path.exists(__file__) else io.StringIO('"'"'from setuptools import setup; setup()'"'"');code = f.read().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, __file__, '"'"'exec'"'"'))' --cpp_ext --cuda_ext install --record /tmp/pip-record-qf8wx_y_/install-record.txt --single-version-externally-managed --compile --install-headers /home/partha9/py38_apex/include/site/python3.8/apex
2021-04-25T17:22:33,327     No CUDA runtime is found, using CUDA_HOME='/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89'

2021-04-25T17:22:33,328     Warning: Torch did not find available GPUs on this system.
2021-04-25T17:22:33,328      If your intention is to cross-compile, this is not an error.
2021-04-25T17:22:33,328     By default, Apex will cross-compile for Pascal (compute capabilities 6.0, 6.1, 6.2),
2021-04-25T17:22:33,328     Volta (compute capability 7.0), Turing (compute capability 7.5),
2021-04-25T17:22:33,328     and, if the CUDA version is >= 11.0, Ampere (compute capability 8.0).
2021-04-25T17:22:33,328     If you wish to cross-compile for a single specific architecture,
2021-04-25T17:22:33,329     export TORCH_CUDA_ARCH_LIST="compute capability" before running setup.py.



2021-04-25T17:22:33,336     torch.__version__  = 1.8.1+cu102


2021-04-25T17:22:33,337     /tmp/pip-req-build-8zq6h56v/setup.py:67: UserWarning: Option --pyprof not specified. Not installing PyProf dependencies!
2021-04-25T17:22:33,337       warnings.warn("Option --pyprof not specified. Not installing PyProf dependencies!")

2021-04-25T17:22:33,344     Compiling cuda extensions with
2021-04-25T17:22:33,344     nvcc: NVIDIA (R) Cuda compiler driver
2021-04-25T17:22:33,344     Copyright (c) 2005-2019 NVIDIA Corporation
2021-04-25T17:22:33,344     Built on Wed_Oct_23_19:24:38_PDT_2019
2021-04-25T17:22:33,344     Cuda compilation tools, release 10.2, V10.2.89
2021-04-25T17:22:33,344     from /cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/bin

2021-04-25T17:22:33,412     running install
2021-04-25T17:22:33,413     running build
2021-04-25T17:22:33,413     running build_ext
2021-04-25T17:22:33,444     building 'apex_C' extension
2021-04-25T17:22:33,444     creating /tmp/pip-req-build-8zq6h56v/build
2021-04-25T17:22:33,444     creating /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8
2021-04-25T17:22:33,444     creating /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc
2021-04-25T17:22:33,467     Emitting ninja build file /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/build.ninja...
2021-04-25T17:22:33,467     Compiling objects...
2021-04-25T17:22:33,467     Using envvar MAX_JOBS (2) as the number of workers...
2021-04-25T17:22:55,650     [1/1] c++ -MMD -MF /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/flatten_unflatten.o.d -Wno-unused-result -Wsign-compare -DNDEBUG -g -fwrapv -O3 -Wall -O2 -ftree-vectorize -march=core-avx2 -fno-math-errno -fPIC -O2 -ftree-vectorize -march=core-avx2 -fno-math-errno -fPIC -fPIC -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/TH -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/THC -I/home/partha9/py38_apex/include -I/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/include/python3.8 -c -c /tmp/pip-req-build-8zq6h56v/csrc/flatten_unflatten.cpp -o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/flatten_unflatten.o -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=apex_C -D_GLIBCXX_USE_CXX11_ABI=0 -std=c++14
2021-04-25T17:22:55,651     In file included from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/Parallel.h:140:0,
2021-04-25T17:22:55,651                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/utils.h:3,
2021-04-25T17:22:55,651                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/nn/cloneable.h:5,
2021-04-25T17:22:55,651                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/nn.h:3,
2021-04-25T17:22:55,651                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/all.h:13,
2021-04-25T17:22:55,651                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/extension.h:4,
2021-04-25T17:22:55,651                      from /tmp/pip-req-build-8zq6h56v/csrc/flatten_unflatten.cpp:1:
2021-04-25T17:22:55,651     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/ParallelOpenMP.h:83:0: warning: ignoring #pragma omp parallel [-Wunknown-pragmas]
2021-04-25T17:22:55,651      #pragma omp parallel for if ((end - begin) >= grain_size)

2021-04-25T17:22:55,651     In file included from /tmp/pip-req-build-8zq6h56v/csrc/flatten_unflatten.cpp:2:0:
2021-04-25T17:22:55,651     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/utils/tensor_flatten.h: In member function ‘at::DeprecatedTypeProperties& torch::utils::TensorGroup::type()’:
2021-04-25T17:22:55,651     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/utils/tensor_flatten.h:44:28: warning: ‘at::DeprecatedTypeProperties& at::Tensor::type() const’ is deprecated: Tensor.type() is deprecated. Instead use Tensor.options(), which in many cases (e.g. in a constructor) is a drop-in replacement. If you were using data from type(), that is now available from Tensor itself, so instead of tensor.type().scalar_type(), use tensor.scalar_type() instead and instead of tensor.type().backend() use tensor.device(). [-Wdeprecated-declarations]
2021-04-25T17:22:55,651          return tensors[0].type();
2021-04-25T17:22:55,651                                 ^
2021-04-25T17:22:55,652     In file included from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/Tensor.h:3:0,
2021-04-25T17:22:55,652                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/Context.h:4,
2021-04-25T17:22:55,652                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/ATen.h:9,
2021-04-25T17:22:55,652                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/types.h:3,
2021-04-25T17:22:55,652                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader_options.h:4,
2021-04-25T17:22:55,652                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/base.h:3,
2021-04-25T17:22:55,652                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/stateful.h:3,
2021-04-25T17:22:55,652                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader.h:3,
2021-04-25T17:22:55,652                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data.h:3,
2021-04-25T17:22:55,652                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/all.h:8,
2021-04-25T17:22:55,652                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/extension.h:4,
2021-04-25T17:22:55,652                      from /tmp/pip-req-build-8zq6h56v/csrc/flatten_unflatten.cpp:1:
2021-04-25T17:22:55,652     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/core/TensorBody.h:303:30: note: declared here
2021-04-25T17:22:55,652        DeprecatedTypeProperties & type() const {
2021-04-25T17:22:55,652                                   ^~~~
2021-04-25T17:22:55,653     creating build/lib.linux-x86_64-3.8
2021-04-25T17:22:55,653     g++ -shared -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/libffi/3.3/lib64 -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/libffi/3.3/lib -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/gcccore/9.3.0/lib64 -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/gcccore/9.3.0/lib -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/libffi/3.3/lib64 -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/libffi/3.3/lib -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/gcccore/9.3.0/lib64 -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/gcccore/9.3.0/lib /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/flatten_unflatten.o -L/home/partha9/py38_apex/lib/python3.8/site-packages/torch/lib -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/lib -lc10 -ltorch -ltorch_cpu -ltorch_python -o build/lib.linux-x86_64-3.8/apex_C.cpython-38-x86_64-linux-gnu.so
2021-04-25T17:22:56,233     building 'amp_C' extension
2021-04-25T17:22:56,265     Emitting ninja build file /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/build.ninja...
2021-04-25T17:22:56,266     Compiling objects...
2021-04-25T17:22:56,266     Using envvar MAX_JOBS (2) as the number of workers...
2021-04-25T17:23:20,614     [1/11] c++ -MMD -MF /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/amp_C_frontend.o.d -Wno-unused-result -Wsign-compare -DNDEBUG -g -fwrapv -O3 -Wall -O2 -ftree-vectorize -march=core-avx2 -fno-math-errno -fPIC -O2 -ftree-vectorize -march=core-avx2 -fno-math-errno -fPIC -fPIC -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/TH -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/THC -I/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/include -I/home/partha9/py38_apex/include -I/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/include/python3.8 -c -c /tmp/pip-req-build-8zq6h56v/csrc/amp_C_frontend.cpp -o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/amp_C_frontend.o -O3 -DVERSION_GE_1_1 -DVERSION_GE_1_3 -DVERSION_GE_1_5 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=amp_C -D_GLIBCXX_USE_CXX11_ABI=0 -std=c++14
2021-04-25T17:23:20,615     In file included from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/Parallel.h:140:0,
2021-04-25T17:23:20,615                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/utils.h:3,
2021-04-25T17:23:20,615                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/nn/cloneable.h:5,
2021-04-25T17:23:20,615                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/nn.h:3,
2021-04-25T17:23:20,615                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/all.h:13,
2021-04-25T17:23:20,615                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/extension.h:4,
2021-04-25T17:23:20,615                      from /tmp/pip-req-build-8zq6h56v/csrc/amp_C_frontend.cpp:1:
2021-04-25T17:23:20,615     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/ParallelOpenMP.h:83:0: warning: ignoring #pragma omp parallel [-Wunknown-pragmas]
2021-04-25T17:23:20,615      #pragma omp parallel for if ((end - begin) >= grain_size)

2021-04-25T17:23:38,331     [2/11] /cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/bin/nvcc --generate-dependencies-with-compile --dependency-output /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_scale_kernel.o.d -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/TH -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/THC -I/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/include -I/home/partha9/py38_apex/include -I/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/include/python3.8 -c -c /tmp/pip-req-build-8zq6h56v/csrc/multi_tensor_scale_kernel.cu -o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_scale_kernel.o -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -lineinfo -O3 --use_fast_math -DVERSION_GE_1_1 -DVERSION_GE_1_3 -DVERSION_GE_1_5 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=amp_C -D_GLIBCXX_USE_CXX11_ABI=0 -gencode=arch=compute_60,code=sm_60 -gencode=arch=compute_61,code=sm_61 -gencode=arch=compute_62,code=sm_62 -gencode=arch=compute_70,code=sm_70 -gencode=arch=compute_75,code=sm_75 -std=c++14
2021-04-25T17:24:01,187     [3/11] /cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/bin/nvcc --generate-dependencies-with-compile --dependency-output /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_sgd_kernel.o.d -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/TH -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/THC -I/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/include -I/home/partha9/py38_apex/include -I/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/include/python3.8 -c -c /tmp/pip-req-build-8zq6h56v/csrc/multi_tensor_sgd_kernel.cu -o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_sgd_kernel.o -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -lineinfo -O3 --use_fast_math -DVERSION_GE_1_1 -DVERSION_GE_1_3 -DVERSION_GE_1_5 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=amp_C -D_GLIBCXX_USE_CXX11_ABI=0 -gencode=arch=compute_60,code=sm_60 -gencode=arch=compute_61,code=sm_61 -gencode=arch=compute_62,code=sm_62 -gencode=arch=compute_70,code=sm_70 -gencode=arch=compute_75,code=sm_75 -std=c++14
2021-04-25T17:24:22,245     [4/11] /cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/bin/nvcc --generate-dependencies-with-compile --dependency-output /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_axpby_kernel.o.d -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/TH -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/THC -I/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/include -I/home/partha9/py38_apex/include -I/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/include/python3.8 -c -c /tmp/pip-req-build-8zq6h56v/csrc/multi_tensor_axpby_kernel.cu -o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_axpby_kernel.o -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -lineinfo -O3 --use_fast_math -DVERSION_GE_1_1 -DVERSION_GE_1_3 -DVERSION_GE_1_5 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=amp_C -D_GLIBCXX_USE_CXX11_ABI=0 -gencode=arch=compute_60,code=sm_60 -gencode=arch=compute_61,code=sm_61 -gencode=arch=compute_62,code=sm_62 -gencode=arch=compute_70,code=sm_70 -gencode=arch=compute_75,code=sm_75 -std=c++14
2021-04-25T17:24:44,653     [5/11] /cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/bin/nvcc --generate-dependencies-with-compile --dependency-output /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_l2norm_kernel.o.d -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/TH -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/THC -I/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/include -I/home/partha9/py38_apex/include -I/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/include/python3.8 -c -c /tmp/pip-req-build-8zq6h56v/csrc/multi_tensor_l2norm_kernel.cu -o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_l2norm_kernel.o -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -lineinfo -O3 --use_fast_math -DVERSION_GE_1_1 -DVERSION_GE_1_3 -DVERSION_GE_1_5 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=amp_C -D_GLIBCXX_USE_CXX11_ABI=0 -gencode=arch=compute_60,code=sm_60 -gencode=arch=compute_61,code=sm_61 -gencode=arch=compute_62,code=sm_62 -gencode=arch=compute_70,code=sm_70 -gencode=arch=compute_75,code=sm_75 -std=c++14
2021-04-25T17:25:05,723     [6/11] /cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/bin/nvcc --generate-dependencies-with-compile --dependency-output /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_lamb_stage_1.o.d -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/TH -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/THC -I/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/include -I/home/partha9/py38_apex/include -I/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/include/python3.8 -c -c /tmp/pip-req-build-8zq6h56v/csrc/multi_tensor_lamb_stage_1.cu -o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_lamb_stage_1.o -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -lineinfo -O3 --use_fast_math -DVERSION_GE_1_1 -DVERSION_GE_1_3 -DVERSION_GE_1_5 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=amp_C -D_GLIBCXX_USE_CXX11_ABI=0 -gencode=arch=compute_60,code=sm_60 -gencode=arch=compute_61,code=sm_61 -gencode=arch=compute_62,code=sm_62 -gencode=arch=compute_70,code=sm_70 -gencode=arch=compute_75,code=sm_75 -std=c++14
2021-04-25T17:25:26,412     [7/11] /cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/bin/nvcc --generate-dependencies-with-compile --dependency-output /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_lamb_stage_2.o.d -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/TH -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/THC -I/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/include -I/home/partha9/py38_apex/include -I/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/include/python3.8 -c -c /tmp/pip-req-build-8zq6h56v/csrc/multi_tensor_lamb_stage_2.cu -o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_lamb_stage_2.o -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -lineinfo -O3 --use_fast_math -DVERSION_GE_1_1 -DVERSION_GE_1_3 -DVERSION_GE_1_5 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=amp_C -D_GLIBCXX_USE_CXX11_ABI=0 -gencode=arch=compute_60,code=sm_60 -gencode=arch=compute_61,code=sm_61 -gencode=arch=compute_62,code=sm_62 -gencode=arch=compute_70,code=sm_70 -gencode=arch=compute_75,code=sm_75 -std=c++14
2021-04-25T17:25:49,576     [8/11] /cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/bin/nvcc --generate-dependencies-with-compile --dependency-output /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_adam.o.d -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/TH -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/THC -I/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/include -I/home/partha9/py38_apex/include -I/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/include/python3.8 -c -c /tmp/pip-req-build-8zq6h56v/csrc/multi_tensor_adam.cu -o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_adam.o -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -lineinfo -O3 --use_fast_math -DVERSION_GE_1_1 -DVERSION_GE_1_3 -DVERSION_GE_1_5 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=amp_C -D_GLIBCXX_USE_CXX11_ABI=0 -gencode=arch=compute_60,code=sm_60 -gencode=arch=compute_61,code=sm_61 -gencode=arch=compute_62,code=sm_62 -gencode=arch=compute_70,code=sm_70 -gencode=arch=compute_75,code=sm_75 -std=c++14
2021-04-25T17:26:09,328     [9/11] /cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/bin/nvcc --generate-dependencies-with-compile --dependency-output /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_adagrad.o.d -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/TH -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/THC -I/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/include -I/home/partha9/py38_apex/include -I/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/include/python3.8 -c -c /tmp/pip-req-build-8zq6h56v/csrc/multi_tensor_adagrad.cu -o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_adagrad.o -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -lineinfo -O3 --use_fast_math -DVERSION_GE_1_1 -DVERSION_GE_1_3 -DVERSION_GE_1_5 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=amp_C -D_GLIBCXX_USE_CXX11_ABI=0 -gencode=arch=compute_60,code=sm_60 -gencode=arch=compute_61,code=sm_61 -gencode=arch=compute_62,code=sm_62 -gencode=arch=compute_70,code=sm_70 -gencode=arch=compute_75,code=sm_75 -std=c++14
2021-04-25T17:26:35,226     [10/11] /cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/bin/nvcc --generate-dependencies-with-compile --dependency-output /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_novograd.o.d -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/TH -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/THC -I/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/include -I/home/partha9/py38_apex/include -I/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/include/python3.8 -c -c /tmp/pip-req-build-8zq6h56v/csrc/multi_tensor_novograd.cu -o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_novograd.o -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -lineinfo -O3 --use_fast_math -DVERSION_GE_1_1 -DVERSION_GE_1_3 -DVERSION_GE_1_5 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=amp_C -D_GLIBCXX_USE_CXX11_ABI=0 -gencode=arch=compute_60,code=sm_60 -gencode=arch=compute_61,code=sm_61 -gencode=arch=compute_62,code=sm_62 -gencode=arch=compute_70,code=sm_70 -gencode=arch=compute_75,code=sm_75 -std=c++14
2021-04-25T17:26:55,515     [11/11] /cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/bin/nvcc --generate-dependencies-with-compile --dependency-output /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_lamb.o.d -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/TH -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/THC -I/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/include -I/home/partha9/py38_apex/include -I/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/include/python3.8 -c -c /tmp/pip-req-build-8zq6h56v/csrc/multi_tensor_lamb.cu -o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_lamb.o -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -lineinfo -O3 --use_fast_math -DVERSION_GE_1_1 -DVERSION_GE_1_3 -DVERSION_GE_1_5 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=amp_C -D_GLIBCXX_USE_CXX11_ABI=0 -gencode=arch=compute_60,code=sm_60 -gencode=arch=compute_61,code=sm_61 -gencode=arch=compute_62,code=sm_62 -gencode=arch=compute_70,code=sm_70 -gencode=arch=compute_75,code=sm_75 -std=c++14
2021-04-25T17:26:55,515     g++ -shared -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/libffi/3.3/lib64 -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/libffi/3.3/lib -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/gcccore/9.3.0/lib64 -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/gcccore/9.3.0/lib -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/libffi/3.3/lib64 -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/libffi/3.3/lib -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/gcccore/9.3.0/lib64 -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/gcccore/9.3.0/lib /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/amp_C_frontend.o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_sgd_kernel.o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_scale_kernel.o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_axpby_kernel.o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_l2norm_kernel.o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_lamb_stage_1.o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_lamb_stage_2.o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_adam.o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_adagrad.o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_novograd.o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/multi_tensor_lamb.o -L/home/partha9/py38_apex/lib/python3.8/site-packages/torch/lib -L/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/lib64 -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/lib -lc10 -ltorch -ltorch_cpu -ltorch_python -lcudart -lc10_cuda -ltorch_cuda -o build/lib.linux-x86_64-3.8/amp_C.cpython-38-x86_64-linux-gnu.so
2021-04-25T17:26:56,151     building 'syncbn' extension
2021-04-25T17:26:56,180     Emitting ninja build file /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/build.ninja...
2021-04-25T17:26:56,180     Compiling objects...
2021-04-25T17:26:56,180     Using envvar MAX_JOBS (2) as the number of workers...
2021-04-25T17:27:24,039     [1/2] c++ -MMD -MF /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/syncbn.o.d -Wno-unused-result -Wsign-compare -DNDEBUG -g -fwrapv -O3 -Wall -O2 -ftree-vectorize -march=core-avx2 -fno-math-errno -fPIC -O2 -ftree-vectorize -march=core-avx2 -fno-math-errno -fPIC -fPIC -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/TH -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/THC -I/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/include -I/home/partha9/py38_apex/include -I/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/include/python3.8 -c -c /tmp/pip-req-build-8zq6h56v/csrc/syncbn.cpp -o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/syncbn.o -O3 -DVERSION_GE_1_1 -DVERSION_GE_1_3 -DVERSION_GE_1_5 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=syncbn -D_GLIBCXX_USE_CXX11_ABI=0 -std=c++14
2021-04-25T17:27:24,040     In file included from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/Parallel.h:140:0,
2021-04-25T17:27:24,040                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/utils.h:3,
2021-04-25T17:27:24,040                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/nn/cloneable.h:5,
2021-04-25T17:27:24,040                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/nn.h:3,
2021-04-25T17:27:24,040                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/all.h:13,
2021-04-25T17:27:24,040                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/extension.h:4,
2021-04-25T17:27:24,041                      from /tmp/pip-req-build-8zq6h56v/csrc/syncbn.cpp:1:
2021-04-25T17:27:24,041     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/ParallelOpenMP.h:83:0: warning: ignoring #pragma omp parallel [-Wunknown-pragmas]
2021-04-25T17:27:24,041      #pragma omp parallel for if ((end - begin) >= grain_size)

2021-04-25T17:27:44,376     [2/2] /cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/bin/nvcc --generate-dependencies-with-compile --dependency-output /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/welford.o.d -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/TH -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/THC -I/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/include -I/home/partha9/py38_apex/include -I/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/include/python3.8 -c -c /tmp/pip-req-build-8zq6h56v/csrc/welford.cu -o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/welford.o -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -O3 -DVERSION_GE_1_1 -DVERSION_GE_1_3 -DVERSION_GE_1_5 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=syncbn -D_GLIBCXX_USE_CXX11_ABI=0 -gencode=arch=compute_60,code=sm_60 -gencode=arch=compute_61,code=sm_61 -gencode=arch=compute_62,code=sm_62 -gencode=arch=compute_70,code=sm_70 -gencode=arch=compute_75,code=sm_75 -std=c++14
2021-04-25T17:27:44,379     g++ -shared -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/libffi/3.3/lib64 -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/libffi/3.3/lib -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/gcccore/9.3.0/lib64 -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/gcccore/9.3.0/lib -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/libffi/3.3/lib64 -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/libffi/3.3/lib -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/gcccore/9.3.0/lib64 -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/Core/gcccore/9.3.0/lib /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/syncbn.o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/welford.o -L/home/partha9/py38_apex/lib/python3.8/site-packages/torch/lib -L/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/lib64 -L/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/lib -lc10 -ltorch -ltorch_cpu -ltorch_python -lcudart -lc10_cuda -ltorch_cuda -o build/lib.linux-x86_64-3.8/syncbn.cpython-38-x86_64-linux-gnu.so
2021-04-25T17:27:45,002     building 'fused_layer_norm_cuda' extension
2021-04-25T17:27:45,027     Emitting ninja build file /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/build.ninja...
2021-04-25T17:27:45,027     Compiling objects...
2021-04-25T17:27:45,027     Using envvar MAX_JOBS (2) as the number of workers...
2021-04-25T17:28:09,036     [1/2] c++ -MMD -MF /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/layer_norm_cuda.o.d -Wno-unused-result -Wsign-compare -DNDEBUG -g -fwrapv -O3 -Wall -O2 -ftree-vectorize -march=core-avx2 -fno-math-errno -fPIC -O2 -ftree-vectorize -march=core-avx2 -fno-math-errno -fPIC -fPIC -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/TH -I/home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/THC -I/cvmfs/soft.computecanada.ca/easybuild/software/2017/Core/cudacore/10.2.89/include -I/home/partha9/py38_apex/include -I/cvmfs/soft.computecanada.ca/easybuild/software/2020/avx2/Core/python/3.8.2/include/python3.8 -c -c /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp -o /tmp/pip-req-build-8zq6h56v/build/temp.linux-x86_64-3.8/csrc/layer_norm_cuda.o -O3 -DVERSION_GE_1_1 -DVERSION_GE_1_3 -DVERSION_GE_1_5 -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=fused_layer_norm_cuda -D_GLIBCXX_USE_CXX11_ABI=0 -std=c++14
2021-04-25T17:28:09,037     In file included from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/Parallel.h:140:0,
2021-04-25T17:28:09,037                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/utils.h:3,
2021-04-25T17:28:09,037                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/nn/cloneable.h:5,
2021-04-25T17:28:09,037                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/nn.h:3,
2021-04-25T17:28:09,038                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/all.h:13,
2021-04-25T17:28:09,038                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/extension.h:4,
2021-04-25T17:28:09,038                      from /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:1:
2021-04-25T17:28:09,038     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/ParallelOpenMP.h:83:0: warning: ignoring #pragma omp parallel [-Wunknown-pragmas]
2021-04-25T17:28:09,038      #pragma omp parallel for if ((end - begin) >= grain_size)

2021-04-25T17:28:09,038     In file included from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/core/DeviceType.h:8:0,
2021-04-25T17:28:09,038                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/core/Device.h:3,
2021-04-25T17:28:09,038                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/core/Allocator.h:6,
2021-04-25T17:28:09,038                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/ATen.h:7,
2021-04-25T17:28:09,038                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/types.h:3,
2021-04-25T17:28:09,038                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader_options.h:4,
2021-04-25T17:28:09,038                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/base.h:3,
2021-04-25T17:28:09,038                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/stateful.h:3,
2021-04-25T17:28:09,038                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader.h:3,
2021-04-25T17:28:09,038                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data.h:3,
2021-04-25T17:28:09,038                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/all.h:8,
2021-04-25T17:28:09,038                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/extension.h:4,
2021-04-25T17:28:09,039                      from /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:1:
2021-04-25T17:28:09,039     /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp: In function ‘std::vector<at::Tensor> layer_norm(at::Tensor, c10::IntArrayRef, double)’:
2021-04-25T17:28:09,039     /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:117:42: warning: ‘at::DeprecatedTypeProperties& at::Tensor::type() const’ is deprecated: Tensor.type() is deprecated. Instead use Tensor.options(), which in many cases (e.g. in a constructor) is a drop-in replacement. If you were using data from type(), that is now available from Tensor itself, so instead of tensor.type().scalar_type(), use tensor.scalar_type() instead and instead of tensor.type().backend() use tensor.device(). [-Wdeprecated-declarations]
2021-04-25T17:28:09,039      #define CHECK_CUDA(x) TORCH_CHECK(x.type().is_cuda(), #x " must be a CUDA tensor")
2021-04-25T17:28:09,039                                               ^
2021-04-25T17:28:09,039     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/macros/Macros.h:173:65: note: in definition of macro ‘C10_UNLIKELY’
2021-04-25T17:28:09,039      #define C10_UNLIKELY(expr)  (__builtin_expect(static_cast<bool>(expr), 0))
2021-04-25T17:28:09,039                                                                      ^~~~
2021-04-25T17:28:09,039     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/util/Exception.h:362:7: note: in expansion of macro ‘C10_UNLIKELY_OR_CONST’
2021-04-25T17:28:09,039        if (C10_UNLIKELY_OR_CONST(!(cond))) {                                 \
2021-04-25T17:28:09,039            ^~~~~~~~~~~~~~~~~~~~~
2021-04-25T17:28:09,039     /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:117:23: note: in expansion of macro ‘TORCH_CHECK’
2021-04-25T17:28:09,039      #define CHECK_CUDA(x) TORCH_CHECK(x.type().is_cuda(), #x " must be a CUDA tensor")
2021-04-25T17:28:09,039                            ^~~~~~~~~~~
2021-04-25T17:28:09,039     /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:119:24: note: in expansion of macro ‘CHECK_CUDA’
2021-04-25T17:28:09,040      #define CHECK_INPUT(x) CHECK_CUDA(x); CHECK_CONTIGUOUS(x)
2021-04-25T17:28:09,040                             ^~~~~~~~~~
2021-04-25T17:28:09,040     /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:129:3: note: in expansion of macro ‘CHECK_INPUT’
2021-04-25T17:28:09,040        CHECK_INPUT(input);
2021-04-25T17:28:09,040        ^
2021-04-25T17:28:09,040     In file included from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/Tensor.h:3:0,
2021-04-25T17:28:09,040                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/Context.h:4,
2021-04-25T17:28:09,040                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/ATen.h:9,
2021-04-25T17:28:09,040                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/types.h:3,
2021-04-25T17:28:09,040                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader_options.h:4,
2021-04-25T17:28:09,040                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/base.h:3,
2021-04-25T17:28:09,040                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/stateful.h:3,
2021-04-25T17:28:09,040                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader.h:3,
2021-04-25T17:28:09,040                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data.h:3,
2021-04-25T17:28:09,040                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/all.h:8,
2021-04-25T17:28:09,040                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/extension.h:4,
2021-04-25T17:28:09,040                      from /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:1:
2021-04-25T17:28:09,040     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/core/TensorBody.h:303:30: note: declared here
2021-04-25T17:28:09,040        DeprecatedTypeProperties & type() const {
2021-04-25T17:28:09,041                                   ^~~~
2021-04-25T17:28:09,041     In file included from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/core/DeviceType.h:8:0,
2021-04-25T17:28:09,041                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/core/Device.h:3,
2021-04-25T17:28:09,041                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/core/Allocator.h:6,
2021-04-25T17:28:09,041                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/ATen.h:7,
2021-04-25T17:28:09,041                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/types.h:3,
2021-04-25T17:28:09,041                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader_options.h:4,
2021-04-25T17:28:09,041                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/base.h:3,
2021-04-25T17:28:09,041                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/stateful.h:3,
2021-04-25T17:28:09,041                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader.h:3,
2021-04-25T17:28:09,041                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data.h:3,
2021-04-25T17:28:09,041                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/all.h:8,
2021-04-25T17:28:09,041                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/extension.h:4,
2021-04-25T17:28:09,041                      from /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:1:
2021-04-25T17:28:09,041     /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp: In function ‘std::vector<at::Tensor> layer_norm_affine(at::Tensor, c10::IntArrayRef, at::Tensor, at::Tensor, double)’:
2021-04-25T17:28:09,041     /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:117:42: warning: ‘at::DeprecatedTypeProperties& at::Tensor::type() const’ is deprecated: Tensor.type() is deprecated. Instead use Tensor.options(), which in many cases (e.g. in a constructor) is a drop-in replacement. If you were using data from type(), that is now available from Tensor itself, so instead of tensor.type().scalar_type(), use tensor.scalar_type() instead and instead of tensor.type().backend() use tensor.device(). [-Wdeprecated-declarations]
2021-04-25T17:28:09,041      #define CHECK_CUDA(x) TORCH_CHECK(x.type().is_cuda(), #x " must be a CUDA tensor")
2021-04-25T17:28:09,042                                               ^
2021-04-25T17:28:09,042     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/macros/Macros.h:173:65: note: in definition of macro ‘C10_UNLIKELY’
2021-04-25T17:28:09,042      #define C10_UNLIKELY(expr)  (__builtin_expect(static_cast<bool>(expr), 0))
2021-04-25T17:28:09,042                                                                      ^~~~
2021-04-25T17:28:09,042     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/util/Exception.h:362:7: note: in expansion of macro ‘C10_UNLIKELY_OR_CONST’
2021-04-25T17:28:09,042        if (C10_UNLIKELY_OR_CONST(!(cond))) {                                 \
2021-04-25T17:28:09,042            ^~~~~~~~~~~~~~~~~~~~~
2021-04-25T17:28:09,042     /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:117:23: note: in expansion of macro ‘TORCH_CHECK’
2021-04-25T17:28:09,042      #define CHECK_CUDA(x) TORCH_CHECK(x.type().is_cuda(), #x " must be a CUDA tensor")
2021-04-25T17:28:09,042                            ^~~~~~~~~~~
2021-04-25T17:28:09,042     /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:119:24: note: in expansion of macro ‘CHECK_CUDA’
2021-04-25T17:28:09,042      #define CHECK_INPUT(x) CHECK_CUDA(x); CHECK_CONTIGUOUS(x)
2021-04-25T17:28:09,042                             ^~~~~~~~~~
2021-04-25T17:28:09,042     /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:149:3: note: in expansion of macro ‘CHECK_INPUT’
2021-04-25T17:28:09,042        CHECK_INPUT(input);
2021-04-25T17:28:09,042        ^
2021-04-25T17:28:09,042     In file included from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/Tensor.h:3:0,
2021-04-25T17:28:09,043                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/Context.h:4,
2021-04-25T17:28:09,043                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/ATen.h:9,
2021-04-25T17:28:09,043                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/types.h:3,
2021-04-25T17:28:09,043                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader_options.h:4,
2021-04-25T17:28:09,043                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/base.h:3,
2021-04-25T17:28:09,043                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/stateful.h:3,
2021-04-25T17:28:09,043                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader.h:3,
2021-04-25T17:28:09,043                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data.h:3,
2021-04-25T17:28:09,043                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/all.h:8,
2021-04-25T17:28:09,043                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/extension.h:4,
2021-04-25T17:28:09,043                      from /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:1:
2021-04-25T17:28:09,043     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/core/TensorBody.h:303:30: note: declared here
2021-04-25T17:28:09,043        DeprecatedTypeProperties & type() const {
2021-04-25T17:28:09,043                                   ^~~~
2021-04-25T17:28:09,043     In file included from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/core/DeviceType.h:8:0,
2021-04-25T17:28:09,043                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/core/Device.h:3,
2021-04-25T17:28:09,044                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/core/Allocator.h:6,
2021-04-25T17:28:09,044                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/ATen.h:7,
2021-04-25T17:28:09,044                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/types.h:3,
2021-04-25T17:28:09,044                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader_options.h:4,
2021-04-25T17:28:09,044                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/base.h:3,
2021-04-25T17:28:09,044                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/stateful.h:3,
2021-04-25T17:28:09,044                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader.h:3,
2021-04-25T17:28:09,044                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data.h:3,
2021-04-25T17:28:09,044                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/all.h:8,
2021-04-25T17:28:09,044                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/extension.h:4,
2021-04-25T17:28:09,044                      from /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:1:
2021-04-25T17:28:09,044     /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:117:42: warning: ‘at::DeprecatedTypeProperties& at::Tensor::type() const’ is deprecated: Tensor.type() is deprecated. Instead use Tensor.options(), which in many cases (e.g. in a constructor) is a drop-in replacement. If you were using data from type(), that is now available from Tensor itself, so instead of tensor.type().scalar_type(), use tensor.scalar_type() instead and instead of tensor.type().backend() use tensor.device(). [-Wdeprecated-declarations]
2021-04-25T17:28:09,044      #define CHECK_CUDA(x) TORCH_CHECK(x.type().is_cuda(), #x " must be a CUDA tensor")
2021-04-25T17:28:09,044                                               ^
2021-04-25T17:28:09,044     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/macros/Macros.h:173:65: note: in definition of macro ‘C10_UNLIKELY’
2021-04-25T17:28:09,044      #define C10_UNLIKELY(expr)  (__builtin_expect(static_cast<bool>(expr), 0))
2021-04-25T17:28:09,045                                                                      ^~~~
2021-04-25T17:28:09,045     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/util/Exception.h:362:7: note: in expansion of macro ‘C10_UNLIKELY_OR_CONST’
2021-04-25T17:28:09,045        if (C10_UNLIKELY_OR_CONST(!(cond))) {                                 \
2021-04-25T17:28:09,045            ^~~~~~~~~~~~~~~~~~~~~
2021-04-25T17:28:09,045     /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:117:23: note: in expansion of macro ‘TORCH_CHECK’
2021-04-25T17:28:09,045      #define CHECK_CUDA(x) TORCH_CHECK(x.type().is_cuda(), #x " must be a CUDA tensor")
2021-04-25T17:28:09,045                            ^~~~~~~~~~~
2021-04-25T17:28:09,045     /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:119:24: note: in expansion of macro ‘CHECK_CUDA’
2021-04-25T17:28:09,045      #define CHECK_INPUT(x) CHECK_CUDA(x); CHECK_CONTIGUOUS(x)
2021-04-25T17:28:09,045                             ^~~~~~~~~~
2021-04-25T17:28:09,045     /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:150:3: note: in expansion of macro ‘CHECK_INPUT’
2021-04-25T17:28:09,045        CHECK_INPUT(gamma);
2021-04-25T17:28:09,045        ^
2021-04-25T17:28:09,045     In file included from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/Tensor.h:3:0,
2021-04-25T17:28:09,045                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/Context.h:4,
2021-04-25T17:28:09,045                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/ATen.h:9,
2021-04-25T17:28:09,046                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/types.h:3,
2021-04-25T17:28:09,046                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader_options.h:4,
2021-04-25T17:28:09,046                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/base.h:3,
2021-04-25T17:28:09,046                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/stateful.h:3,
2021-04-25T17:28:09,046                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader.h:3,
2021-04-25T17:28:09,046                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data.h:3,
2021-04-25T17:28:09,046                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/all.h:8,
2021-04-25T17:28:09,046                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/extension.h:4,
2021-04-25T17:28:09,046                      from /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:1:
2021-04-25T17:28:09,046     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/core/TensorBody.h:303:30: note: declared here
2021-04-25T17:28:09,046        DeprecatedTypeProperties & type() const {
2021-04-25T17:28:09,046                                   ^~~~
2021-04-25T17:28:09,046     In file included from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/core/DeviceType.h:8:0,
2021-04-25T17:28:09,046                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/core/Device.h:3,
2021-04-25T17:28:09,046                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/core/Allocator.h:6,
2021-04-25T17:28:09,047                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/ATen/ATen.h:7,
2021-04-25T17:28:09,047                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/types.h:3,
2021-04-25T17:28:09,047                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader_options.h:4,
2021-04-25T17:28:09,047                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/base.h:3,
2021-04-25T17:28:09,047                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/stateful.h:3,
2021-04-25T17:28:09,047                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader.h:3,
2021-04-25T17:28:09,047                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/data.h:3,
2021-04-25T17:28:09,047                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/csrc/api/include/torch/all.h:8,
2021-04-25T17:28:09,047                      from /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/torch/extension.h:4,
2021-04-25T17:28:09,047                      from /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:1:
2021-04-25T17:28:09,047     /tmp/pip-req-build-8zq6h56v/csrc/layer_norm_cuda.cpp:117:42: warning: ‘at::DeprecatedTypeProperties& at::Tensor::type() const’ is deprecated: Tensor.type() is deprecated. Instead use Tensor.options(), which in many cases (e.g. in a constructor) is a drop-in replacement. If you were using data from type(), that is now available from Tensor itself, so instead of tensor.type().scalar_type(), use tensor.scalar_type() instead and instead of tensor.type().backend() use tensor.device(). [-Wdeprecated-declarations]
2021-04-25T17:28:09,047      #define CHECK_CUDA(x) TORCH_CHECK(x.type().is_cuda(), #x " must be a CUDA tensor")
2021-04-25T17:28:09,047                                               ^
2021-04-25T17:28:09,047     /home/partha9/py38_apex/lib/python3.8/site-packages/torch/include/c10/macros/Macros.h:173:65: note: in definitio

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with setup.py and csrc/flatten_unflatten.cpp, then inspect the full build log around the cublasLtMatmulDescOpaque_t error and the reported PyTorch 1.8.1/CUDA 10.2 environment. Done means identifying whether this version combination is supported and documenting or validating a reproducible resolution for the Apex build.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp, python, pytorch
Domain
build-system, machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.