I am trying to build Pytorch from the source without CUDA and ROCM on linux
OS: CentOS Linux release 8.2.2004 and Ubuntu 18.04.5 LTS
Got error below:
./pytorch/torch/csrc/jit/ir/ir.cpp:1189:16: error: ‘set_stream’ is not a member of ‘torch::jit::cuda’
case cuda::set_stream:
^~~~~~~~~~
./pytorch/torch/csrc/jit/ir/ir.cpp:1189:16: note: suggested alternative:
In file included from./pytorch/torch/csrc/jit/ir/ir.h:18,
from ./pytorch/torch/csrc/jit/ir/ir.cpp:1:
./pytorch/aten/src/ATen/core/interned_strings.h:221:11: note: ‘c10::cuda::set_stream’
_(cuda, set_stream) \
^~~~~~~~~~
./pytorch/aten/src/ATen/core/interned_strings.h:344:35: note: in definition of macro ‘DEFINE_SYMBOL’
namespace ns { constexpr Symbol s(static_cast<unique_t>(_keys::ns##_##s)); }
^
./pytorch/aten/src/ATen/core/interned_strings.h:345:1: note: in expansion of macro ‘FORALL_NS_SYMBOLS’
FORALL_NS_SYMBOLS(DEFINE_SYMBOL)
^~~~~~~~~~~~~~~~~
./pytorch/torch/csrc/jit/ir/ir.cpp:1190:16: error: ‘_set_device’ is not a member of ‘torch::jit::cuda’
case cuda::_set_device:
^~~~~~~~~~~
./pytorch/torch/csrc/jit/ir/ir.cpp:1190:16: note: suggested alternative:
In file included from /proj/gdba/vivianw/pytorch/torch/csrc/jit/ir/ir.h:18,
from /proj/gdba/vivianw/pytorch/torch/csrc/jit/ir/ir.cpp:1:
./pytorch/aten/src/ATen/core/interned_strings.h:220:11: note: ‘c10::cuda::_set_device’
_(cuda, _set_device) \
^~~~~~~~~~~
./pytorch/aten/src/ATen/core/interned_strings.h:344:35: note: in definition of macro ‘DEFINE_SYMBOL’
namespace ns { constexpr Symbol s(static_cast<unique_t>(_keys::ns##_##s)); }
^
./pytorch/aten/src/ATen/core/interned_strings.h:345:1: note: in expansion of macro ‘FORALL_NS_SYMBOLS’
FORALL_NS_SYMBOLS(DEFINE_SYMBOL)
^~~~~~~~~~~~~~~~~
./pytorch/torch/csrc/jit/ir/ir.cpp:1191:16: error: ‘_current_device’ is not a member of ‘torch::jit::cuda’
case cuda::_current_device:
^~~~~~~~~~~~~~~
./pytorch/torch/csrc/jit/ir/ir.cpp:1191:16: note: suggested alternative:
In file included from /proj/gdba/vivianw/pytorch/torch/csrc/jit/ir/ir.h:18,
from /proj/gdba/vivianw/pytorch/torch/csrc/jit/ir/ir.cpp:1:
./pytorch/aten/src/ATen/core/interned_strings.h:222:11: note: ‘c10::cuda::_current_device’
_(cuda, _current_device) \
^~~~~~~~~~~~~~~
.....
[70/849] Building CXX object caffe2/CMakeFiles/torch_cpu.dir/__/torch/csrc/jit/passes/tensorexpr_fuser.cpp.o
ninja: build stopped: subcommand failed.
Steps I have run:
(base) bash-4.4$ conda create --name pytorch-dev python=3.9.12
(base) bash-4.4$ conda activate pytorch-dev
#install dependencies
(pytorch-dev) bash-4.4$ conda install astunparse numpy ninja pyyaml setuptools cmake cffi typing_extensions future six requests dataclasses
(pytorch-dev) bash-4.4$ conda install mkl mkl-include
#Get the PyTorch Source
(pytorch-dev) bash-4.4$git clone --recursive https://github.com/pytorch/pytorch
(pytorch-dev) bash-4.4$cd pytorch
# if you are updating an existing checkout
(pytorch-dev) bash-4.4$git submodule sync
(pytorch-dev) bash-4.4$git submodule update --init --recursive --jobs 0
#Install PyTorch
(pytorch-dev) bash-4.4$ export CMAKE_PREFIX_PATH=${CONDA_PREFIX:-"$(dirname $(which conda))/../"}
(pytorch-dev) bash-4.4$ USE_ROCM=0 USE_CUDA=0 python setup.py install