The current PyTorch install supports CUDA capabilities sm_37 sm_50 sm_60 sm_70.

Question

The current PyTorch install supports CUDA capabilities sm_37 sm_50 sm_60 sm_70.

zhourui-xihu opened this issue 2 years ago · comments

I ran the code on Nvidia a100 gpu and reported the following error:

UserWarning:
NVIDIA A100-SXM4-80GB with CUDA capability sm_80 is not compatible with the curr ent PyTorch installation.
The current PyTorch install supports CUDA capabilities sm_37 sm_50 sm_60 sm_70.
If you want to use the NVIDIA A100-SXM4-80GB GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/

warnings.warn(incompatible_device_warn.format(device_name, capability, " ".joi n(arch_list), device_name))
Traceback (most recent call last):
File "/data/home/zhourui/miniconda3/envs/diff/lib/python3.8/site-packages/torc h/utils/cpp_extension.py", line 1808, in _run_ninja_build
subprocess.run(
File "/data/home/zhourui/miniconda3/envs/diff/lib/python3.8/subprocess.py", li ne 516, in run
raise CalledProcessError(retcode, process.args,
subprocess.CalledProcessError: Command '['ninja', '-v']' returned non-zero exit status 1.

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
File "train.py", line 11, in
from sgmse.backbones.shared import BackboneRegistry
File "/data/home/zhourui/code/sgmse-main/sgmse/backbones/init.py", line 2, in
from .ncsnpp import NCSNpp
File "/data/home/zhourui/code/sgmse-main/sgmse/backbones/ncsnpp.py", line 18, in
from .ncsnpp_utils import layers, layerspp, normalization
File "/data/home/zhourui/code/sgmse-main/sgmse/backbones/ncsnpp_utils/layerspp .py", line 20, in
from . import up_or_down_sampling
File "/data/home/zhourui/code/sgmse-main/sgmse/backbones/ncsnpp_utils/up_or_do wn_sampling.py", line 10, in
from .op import upfirdn2d
File "/data/home/zhourui/code/sgmse-main/sgmse/backbones/ncsnpp_utils/op/ini t.py", line 1, in
from .fused_act import FusedLeakyReLU, fused_leaky_relu
File "/data/home/zhourui/code/sgmse-main/sgmse/backbones/ncsnpp_utils/op/fused act.py", line 11, in
fused = load(
File "/data/home/zhourui/miniconda3/envs/diff/lib/python3.8/site-packages/torc h/utils/cpp_extension.py", line 1202, in load
return jit_compile(
File "/data/home/zhourui/miniconda3/envs/diff/lib/python3.8/site-packages/torc h/utils/cpp_extension.py", line 1425, in jit_compile
write_ninja_file_and_build_library(
File "/data/home/zhourui/miniconda3/envs/diff/lib/python3.8/site-packages/torc h/utils/cpp_extension.py", line 1537, in write_ninja_file_and_build_library
run_ninja_build(
File "/data/home/zhourui/miniconda3/envs/diff/lib/python3.8/site-packages/torc h/utils/cpp_extension.py", line 1824, in run_ninja_build
raise RuntimeError(message) from e
RuntimeError: Error building extension 'fused': [1/2] /usr/local/cuda/bin/nvcc -DTORCH_EXTENSION_NAME=fused -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER TYPE="gcc" -DPYBIND11_STDLIB="libstdcpp" -DPYBIND11_BUILD_ABI="cxxabi10 11" -isystem /data/home/zhourui/miniconda3/envs/diff/lib/python3.8/site-package s/torch/include -isystem /data/home/zhourui/miniconda3/envs/diff/lib/python3.8/s ite-packages/torch/include/torch/csrc/api/include -isystem /data/home/zhourui/mi niconda3/envs/diff/lib/python3.8/site-packages/torch/include/TH -isystem /data/h ome/zhourui/miniconda3/envs/diff/lib/python3.8/site-packages/torch/include/THC - isystem /usr/local/cuda/include -isystem /data/home/zhourui/miniconda3/envs/diff /include/python3.8 -D_GLIBCXX_USE_CXX11_ABI=0 -D__CUDA_NO_HALF_OPERATORS -D__C UDA_NO_HALF_CONVERSIONS -D__CUDA_NO_BFLOAT16_CONVERSIONS -D__CUDA_NO_HALF2_O PERATORS --expt-relaxed-constexpr -gencode=arch=compute_70,code=compute_70 -ge ncode=arch=compute_70,code=sm_70 --compiler-options '-fPIC' -std=c++14 -c /data/ home/zhourui/code/sgmse-main/sgmse/backbones/ncsnpp_utils/op/fused_bias_act_kern el.cu -o fused_bias_act_kernel.cuda.o
FAILED: fused_bias_act_kernel.cuda.o
/usr/local/cuda/bin/nvcc -DTORCH_EXTENSION_NAME=fused -DTORCH_API_INCLUDE_EXTEN SION_H -DPYBIND11_COMPILER_TYPE="gcc" -DPYBIND11_STDLIB="libstdcpp" -DPYBI ND11_BUILD_ABI="cxxabi1011" -isystem /data/home/zhourui/miniconda3/envs/diff/ lib/python3.8/site-packages/torch/include -isystem /data/home/zhourui/miniconda3 /envs/diff/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -isy stem /data/home/zhourui/miniconda3/envs/diff/lib/python3.8/site-packages/torch/i nclude/TH -isystem /data/home/zhourui/miniconda3/envs/diff/lib/python3.8/site-pa ckages/torch/include/THC -isystem /usr/local/cuda/include -isystem /data/home/zh ourui/miniconda3/envs/diff/include/python3.8 -D_GLIBCXX_USE_CXX11_ABI=0 -D__CUDA NO_HALF_OPERATORS -D__CUDA_NO_HALF_CONVERSIONS -D__CUDA_NO_BFLOAT16_CONVERS IONS -D__CUDA_NO_HALF2_OPERATORS_ --expt-relaxed-constexpr -gencode=arch=comp ute_70,code=compute_70 -gencode=arch=compute_70,code=sm_70 --compiler-options '- fPIC' -std=c++14 -c /data/home/zhourui/code/sgmse-main/sgmse/backbones/ncsnpp_ut ils/op/fused_bias_act_kernel.cu -o fused_bias_act_kernel.cuda.o
In file included from /usr/local/cuda/include/cuda_runtime.h:83,
from :
/usr/local/cuda/include/crt/host_config.h:139:2: error: #error -- unsupported GN U version! gcc versions later than 10 are not supported! The nvcc flag '-allow-u nsupported-compiler' can be used to override this version check; however, using an unsupported host compiler may cause compilation failure or incorrect run time execution. Use at your own risk.
139 | #error -- unsupported GNU version! gcc versions later than 10 are not su pported! The nvcc flag '-allow-unsupported-compiler' can be used to override thi s version check; however, using an unsupported host compiler may cause compilati on failure or incorrect run time execution. Use at your own risk.
| ^~~~~
ninja: build stopped: subcommand failed.

k-sparrow · Answer 1 · Sun Jan 01 2023 21:02:19 GMT+0800 (China Standard Time)

Had the same problem with a 3080Ti.

Basically all Ampere cards will get this error.

I upgraded pytorch (also torchvision & torchaudio) version to latest release and it ran just fine.

zhourui-xihu · Answer 2 · Thu Jan 19 2023 15:53:40 GMT+0800 (China Standard Time)

Thank you very much.
Do you encounter this kind of error when testing？

Exception has occurred: RuntimeError
Error building extension 'fused': [1/2] /usr/bin/nvcc -DTORCH_EXTENSION_NAME=fused -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE="gcc" -DPYBIND11_STDLIB="libstdcpp" -DPYBIND11_BUILD_ABI="cxxabi1013" -isystem /data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include -isystem /data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -isystem /data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/TH -isystem /data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/THC -isystem /data/home/zhourui/miniconda3/envs/diffusion/include/python3.8 -D_GLIBCXX_USE_CXX11_ABI=0 -D__CUDA_NO_HALF_OPERATORS -D__CUDA_NO_HALF_CONVERSIONS_ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr -gencode=arch=compute_80,code=compute_80 -gencode=arch=compute_80,code=sm_80 --compiler-options '-fPIC' -std=c++14 -c /data/home/zhourui/code/sgmse-main-rev/sgmse/backbones/ncsnpp_utils/op/fused_bias_act_kernel.cu -o fused_bias_act_kernel.cuda.o
FAILED: fused_bias_act_kernel.cuda.o
/usr/bin/nvcc -DTORCH_EXTENSION_NAME=fused -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE="gcc" -DPYBIND11_STDLIB="libstdcpp" -DPYBIND11_BUILD_ABI="cxxabi1013" -isystem /data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include -isystem /data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -isystem /data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/TH -isystem /data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/THC -isystem /data/home/zhourui/miniconda3/envs/diffusion/include/python3.8 -D_GLIBCXX_USE_CXX11_ABI=0 -D__CUDA_NO_HALF_OPERATORS -D__CUDA_NO_HALF_CONVERSIONS_ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr -gencode=arch=compute_80,code=compute_80 -gencode=arch=compute_80,code=sm_80 --compiler-options '-fPIC' -std=c++14 -c /data/home/zhourui/code/sgmse-main-rev/sgmse/backbones/ncsnpp_utils/op/fused_bias_act_kernel.cu -o fused_bias_act_kernel.cuda.o
/usr/include/c++/11/type_traits(1406): error: type name is not allowed

/usr/include/c++/11/type_traits(1406): error: type name is not allowed

/usr/include/c++/11/type_traits(1406): error: identifier "__is_same" is undefined

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Metaprogramming.h(425): error: no instance of constructor "std::tuple<_T1, _T2>::tuple [with _T1=int32_t &&, _T2=int32_t]" matches the argument list
argument types are: (int32_t, int)

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Metaprogramming.h(423): error: no instance of overloaded function "c10::guts::detail::extract_tuple_element_by_index" matches the argument list
argument types are: (std::tuple<int32_t>, )

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/ScalarTypeToTypeMeta.h(35): error: no suitable constructor exists to convert from "c10::ScalarType" to "c10::optionalc10::ScalarType"

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorOptions.h(148): error: no suitable constructor exists to convert from "c10::Layout" to "c10::optionalc10::Layout"

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorOptions.h(177): error: no instance of overloaded function "c10::TensorOptions::set_dtype" matches the argument list
argument types are: (caffe2::TypeMeta)
object type is: c10::TensorOptions

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorOptions.h(182): error: no instance of overloaded function "c10::TensorOptions::set_dtype" matches the argument list
argument types are: (c10::ScalarType)
object type is: c10::TensorOptions

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorOptions.h(187): error: no suitable constructor exists to convert from "c10::MemoryFormat" to "c10::optionalc10::MemoryFormat"

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(1215): error: no instance of constructor "c10::optional::optional [with T=c10::Device]" matches the argument list
argument types are: (const c10::Device)
detected during instantiation of "c10::optional<std::decay::type> c10::make_optional(T &&) [with T=const c10::Device &]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorOptions.h(287): here

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(1215): error: no instance of constructor "c10::optional::optional [with T=caffe2::TypeMeta]" matches the argument list
argument types are: (const caffe2::TypeMeta)
detected during instantiation of "c10::optional<std::decay::type> c10::make_optional(T &&) [with T=const caffe2::TypeMeta &]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorOptions.h(308): here

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(1215): error: no instance of constructor "c10::optional::optional [with T=c10::Layout]" matches the argument list
argument types are: (const c10::Layout)
detected during instantiation of "c10::optional<std::decay::type> c10::make_optional(T &&) [with T=const c10::Layout &]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorOptions.h(324): here

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(1215): error: no instance of constructor "c10::optional::optional [with T=__nv_bool]" matches the argument list
argument types are: (const __nv_bool)
detected during instantiation of "c10::optional<std::decay::type> c10::make_optional(T &&) [with T=const __nv_bool &]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorOptions.h(341): here

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(1215): error: no instance of constructor "c10::optional::optional [with T=c10::MemoryFormat]" matches the argument list
argument types are: (const c10::MemoryFormat)
detected during instantiation of "c10::optional<std::decay::type> c10::make_optional(T &&) [with T=const c10::MemoryFormat &]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorOptions.h(388): here

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorOptions.h(430): error: no suitable constructor exists to convert from "c10::MemoryFormat" to "c10::optionalc10::MemoryFormat"

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorOptions.h(568): error: no instance of overloaded function "c10::TensorOptions::dtype" matches the argument list
argument types are: (caffe2::TypeMeta)
object type is: c10::TensorOptions

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorOptions.h(573): error: no instance of overloaded function "c10::TensorOptions::dtype" matches the argument list
argument types are: (caffe2::TypeMeta)
object type is: c10::TensorOptions

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorOptions.h(579): error: no instance of overloaded function "c10::TensorOptions::layout" matches the argument list
argument types are: (c10::Layout)
object type is: c10::TensorOptions

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorOptions.h(598): error: no instance of overloaded function "c10::TensorOptions::requires_grad" matches the argument list
argument types are: (__nv_bool)
object type is: c10::TensorOptions

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorOptions.h(604): error: no suitable constructor exists to convert from "c10::MemoryFormat" to "c10::optionalc10::MemoryFormat"

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorOptions.h(828): error: no instance of overloaded function "c10::TensorOptions::layout" matches the argument list
argument types are: (c10::Layout)
object type is: c10::TensorOptions

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(1215): error: no instance of constructor "c10::optional::optional [with T=PyObject *]" matches the argument list
argument types are: (PyObject *)
detected during instantiation of "c10::optional<std::decay::type> c10::make_optional(T &&) [with T=PyObject *]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorImpl.h(1671): here

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(684): error: class "std::enable_if<, c10::optionalc10::Device &>" has no member "type"
detected during instantiation of "c10::optional::operator= [with T=c10::Device]" based on template argument c10::Device
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/TensorImpl.h(1853): here

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/DeprecatedTypeProperties.h(99): error: no instance of overloaded function "c10::TensorOptions::dtype" matches the argument list
argument types are: (caffe2::TypeMeta)
object type is: c10::TensorOptions

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/DeviceGuard.h(31): error: no instance of constructor "c10::impl::InlineDeviceGuard::InlineDeviceGuard [with T=c10::impl::VirtualGuardImpl]" matches the argument list
argument types are: (c10::Device, const c10::impl::DeviceGuardImplInterface *)

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/DeviceGuard.h(50): error: more than one instance of overloaded function "c10::impl::InlineDeviceGuard::reset_device [with T=c10::impl::VirtualGuardImpl]" matches the argument list:
function template "std::enable_if<, void>::type c10::impl::InlineDeviceGuard::reset_device(c10::Device) [with T=c10::impl::VirtualGuardImpl]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/impl/InlineDeviceGuard.h(136): here
function template "std::enable_if<std::is_same<U, c10::impl::VirtualGuardImpl>::value, void>::type c10::impl::InlineDeviceGuard::reset_device(c10::Device, const c10::impl::DeviceGuardImplInterface *) [with T=c10::impl::VirtualGuardImpl]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/impl/InlineDeviceGuard.h(159): here
argument types are: (c10::Device)
object type is: c10::impl::InlineDeviceGuardc10::impl::VirtualGuardImpl

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/impl/InlineDeviceGuard.h(158): error: class "std::enable_if<, void>" has no member "type"
detected during instantiation of "c10::impl::InlineDeviceGuard::reset_device [with T=c10::impl::VirtualGuardImpl]" based on template argument c10::impl::VirtualGuardImpl
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/DeviceGuard.h(57): here

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/DeviceGuard.h(156): error: no instance of overloaded function "c10::impl::InlineOptionalDeviceGuard::reset_device [with T=c10::impl::VirtualGuardImpl]" matches the argument list
argument types are: (c10::Device)
object type is: c10::impl::InlineOptionalDeviceGuardc10::impl::VirtualGuardImpl

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/DeviceGuard.h(163): error: no instance of overloaded function "c10::impl::InlineOptionalDeviceGuard::reset_device [with T=c10::impl::VirtualGuardImpl]" matches the argument list
argument types are: (c10::Device, const c10::impl::DeviceGuardImplInterface *)
object type is: c10::impl::InlineOptionalDeviceGuardc10::impl::VirtualGuardImpl

/usr/include/c++/11/type_traits(2585): error: class "std::enable_if<, c10::complex>" has no member "type"
detected during:
instantiation of type "std::enable_if_t<, c10::complex>"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/complex.h(176): here
instantiation of "c10::complex::complex [with T=float]" based on template argument
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/TypeCast.h(77): here

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/TypeCast.h(77): error: more than one instance of constructor "c10::complex::complex [with T=float]" matches the argument list:
function template "c10::complex::complex(const std::enable_if_t<std::is_same<U, float>::value, c10::complex> &) [with T=float]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/complex.h(170): here
function template "c10::complex::complex(const std::enable_if_t<std::is_same<U, double>::value, c10::complex> &) [with T=float]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/complex.h(175): here
argument types are: (c10::BFloat16)

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/TypeCast.h(86): error: more than one instance of constructor "c10::complex::complex [with T=float]" matches the argument list:
function template "c10::complex::complex(const std::enable_if_t<std::is_same<U, float>::value, c10::complex> &) [with T=float]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/complex.h(170): here
function template "c10::complex::complex(const std::enable_if_t<std::is_same<U, double>::value, c10::complex> &) [with T=float]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/complex.h(175): here
argument types are: (c10::Half)

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/TypeCast.h(98): error: more than one user-defined conversion from "c10::complex" to "c10::complex" applies:
function template "c10::complex::complex(const std::enable_if_t<std::is_same<U, float>::value, c10::complex> &) [with T=float]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/complex.h(170): here
function template "c10::complex::complex(const std::enable_if_t<std::is_same<U, double>::value, c10::complex> &) [with T=float]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/complex.h(175): here

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/Scalar.h(39): error: no instance of constructor "c10::Scalar::Scalar" matches the argument list
argument types are: (uint8_t, __nv_bool)

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/Scalar.h(39): error: no instance of constructor "c10::Scalar::Scalar" matches the argument list
argument types are: (int8_t, __nv_bool)

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/Scalar.h(39): error: no instance of constructor "c10::Scalar::Scalar" matches the argument list
argument types are: (int16_t, __nv_bool)

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/Scalar.h(39): error: no instance of constructor "c10::Scalar::Scalar" matches the argument list
argument types are: (int, __nv_bool)

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/Scalar.h(39): error: no instance of constructor "c10::Scalar::Scalar" matches the argument list
argument types are: (int64_t, __nv_bool)

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=c10::complex, _Up=__nv_bool]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/core/Scalar.h(40): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during:
instantiation of class "std::is_same<_Tp, _Up> [with _Tp=c10::ArrayRef<int64_t>, _Up=c10::nullopt_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(567): here
instantiation of class "c10::optional [with T=c10::ArrayRef<int64_t>]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/OptionalArrayRef.h(208): here
instantiation of class "c10::OptionalArrayRef [with T=int64_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/OptionalArrayRef.h(216): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during:
instantiation of class "std::is_same<_Tp, _Up> [with _Tp=c10::ArrayRef<int64_t>, _Up=c10::in_place_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(570): here
instantiation of class "c10::optional [with T=c10::ArrayRef<int64_t>]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/OptionalArrayRef.h(208): here
instantiation of class "c10::OptionalArrayRef [with T=int64_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/OptionalArrayRef.h(216): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=c10::ArrayRef<int64_t>, _Up=c10::OptionalArrayRef<int64_t>]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/OptionalArrayRef.h(219): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=c10::optional<__nv_bool>, _Up=c10::OptionalArrayRef<int64_t>]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/CheckMemoryFormat.h(9): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=at::NameType, _Up=c10::OptionalArrayRef<int64_t>]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/Dimname.h(20): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=at::NameType, _Up=c10::in_place_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/Dimname.h(20): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=c10::Symbol, _Up=c10::OptionalArrayRef<int64_t>]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/Dimname.h(41): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=c10::Symbol, _Up=c10::in_place_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/Dimname.h(41): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during:
instantiation of class "std::is_same<_Tp, _Up> [with _Tp=at::Dimname, _Up=at::Dimname]"
/usr/include/c++/11/bits/stl_vector.h(401): here
instantiation of class "std::vector<_Tp, _Alloc> [with _Tp=at::Dimname, _Alloc=std::allocatorat::Dimname]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/NamedTensor.h(69): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=at::Dimname *, _Up=at::Dimname *]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/NamedTensor.h(44): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=c10::intrusive_ptr<c10::TensorImpl, c10::UndefinedTensorImpl>, _Up=c10::OptionalArrayRef<int64_t>]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBase.h(209): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=c10::intrusive_ptr<c10::TensorImpl, c10::UndefinedTensorImpl>, _Up=c10::in_place_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBase.h(209): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during:
instantiation of class "std::is_same<_Tp, _Up> [with _Tp=at::DimnameList, _Up=c10::nullopt_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(567): here
instantiation of class "c10::optional [with T=at::DimnameList]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBase.h(230): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during:
instantiation of class "std::is_same<_Tp, _Up> [with _Tp=at::DimnameList, _Up=c10::in_place_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(570): here
instantiation of class "c10::optional [with T=at::DimnameList]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBase.h(230): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=c10::optionalat::DimnameList, _Up=c10::in_place_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBase.h(231): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=c10::optionalat::DimnameList, _Up=c10::optionalat::DimnameList]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBase.h(231): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=const at::Dimname *, _Up=at::Dimname *]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBase.h(235): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=c10::Layout, _Up=c10::OptionalArrayRef<int64_t>]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBase.h(253): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=std::vector<int64_t, std::allocator<int64_t>>, _Up=c10::OptionalArrayRef<int64_t>]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBase.h(256): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=std::vector<int64_t, std::allocator<int64_t>>, _Up=c10::in_place_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBase.h(256): here

/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBase.h(508): error: no instance of overloaded function "c10::TensorOptions::dtype" matches the argument list
argument types are: (caffe2::TypeMeta)
object type is: c10::TensorOptions

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=std::atomic, _Up=c10::OptionalArrayRef<int64_t>]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBase.h(916): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=std::atomic, _Up=c10::in_place_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBase.h(916): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during:
instantiation of class "std::is_same<_Tp, _Up> [with _Tp=at::TensorBase, _Up=c10::nullopt_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(567): here
instantiation of class "c10::optional [with T=at::TensorBase]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBase.h(951): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during:
instantiation of class "std::is_same<_Tp, _Up> [with _Tp=at::TensorBase, _Up=c10::in_place_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(570): here
instantiation of class "c10::optional [with T=at::TensorBase]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBase.h(951): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during:
instantiation of class "std::is_same<_Tp, _Up> [with _Tp=at::TensorList, _Up=c10::nullopt_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(567): here
instantiation of class "c10::optional [with T=at::TensorList]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBody.h(424): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during:
instantiation of class "std::is_same<_Tp, _Up> [with _Tp=at::TensorList, _Up=c10::in_place_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(570): here
instantiation of class "c10::optional [with T=at::TensorList]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBody.h(424): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during instantiation of class "std::is_same<_Tp, _Up> [with _Tp=c10::nullopt_t, _Up=c10::optionalat::TensorList]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBody.h(424): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during:
instantiation of class "std::is_same<_Tp, _Up> [with _Tp=at::Tensor, _Up=c10::nullopt_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(567): here
instantiation of class "c10::optional [with T=at::Tensor]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBody.h(509): here

/usr/include/c++/11/type_traits(1406): error: type name is not allowed
detected during:
instantiation of class "std::is_same<_Tp, _Up> [with _Tp=at::Tensor, _Up=c10::in_place_t]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/c10/util/Optional.h(570): here
instantiation of class "c10::optional [with T=at::Tensor]"
/data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/ATen/core/TensorBody.h(509): here

Error limit reached.
100 errors detected in the compilation of "/data/home/zhourui/code/sgmse-main-rev/sgmse/backbones/ncsnpp_utils/op/fused_bias_act_kernel.cu".
Compilation terminated.
ninja: build stopped: subcommand failed.
subprocess.CalledProcessError: Command '['ninja', '-v']' returned non-zero exit status 1.

The above exception was the direct cause of the following exception:

File "/data/home/zhourui/code/sgmse-main-rev/sgmse/backbones/ncsnpp_utils/op/fused_act.py", line 11, in
fused = load(
File "/data/home/zhourui/code/sgmse-main-rev/sgmse/backbones/ncsnpp_utils/op/init.py", line 1, in
from .fused_act import FusedLeakyReLU, fused_leaky_relu
File "/data/home/zhourui/code/sgmse-main-rev/sgmse/backbones/ncsnpp_utils/up_or_down_sampling.py", line 10, in
from .op import upfirdn2d
File "/data/home/zhourui/code/sgmse-main-rev/sgmse/backbones/ncsnpp_utils/layerspp.py", line 20, in
from . import up_or_down_sampling
File "/data/home/zhourui/code/sgmse-main-rev/sgmse/backbones/ncsnpp.py", line 18, in
from .ncsnpp_utils import layers, layerspp, normalization
File "/data/home/zhourui/code/sgmse-main-rev/sgmse/backbones/init.py", line 2, in
from .ncsnpp import NCSNpp
File "/data/home/zhourui/code/sgmse-main-rev/sgmse/model.py", line 11, in
from sgmse.backbones import BackboneRegistry
File "/data/home/zhourui/code/sgmse-main-rev/enhancement.py", line 10, in
from sgmse.model import ScoreModel
RuntimeError: Error building extension 'fused': [1/2] /usr/bin/nvcc -DTORCH_EXTENSION_NAME=fused -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE="gcc" -DPYBIND11_STDLIB="libstdcpp" -DPYBIND11_BUILD_ABI="cxxabi1013" -isystem /data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include -isystem /data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -isystem /data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/TH -isystem /data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/THC -isystem /data/home/zhourui/miniconda3/envs/diffusion/include/python3.8 -D_GLIBCXX_USE_CXX11_ABI=0 -D__CUDA_NO_HALF_OPERATORS -D__CUDA_NO_HALF_CONVERSIONS_ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr -gencode=arch=compute_80,code=compute_80 -gencode=arch=compute_80,code=sm_80 --compiler-options '-fPIC' -std=c++14 -c /data/home/zhourui/code/sgmse-main-rev/sgmse/backbones/ncsnpp_utils/op/fused_bias_act_kernel.cu -o fused_bias_act_kernel.cuda.o
FAILED: fused_bias_act_kernel.cuda.o
/usr/bin/nvcc -DTORCH_EXTENSION_NAME=fused -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE="gcc" -DPYBIND11_STDLIB="libstdcpp" -DPYBIND11_BUILD_ABI="cxxabi1013" -isystem /data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include -isystem /data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -isystem /data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/TH -isystem /data/home/zhourui/miniconda3/envs/diffusion/lib/python3.8/site-packages/torch/include/THC -isystem /data/home/zhourui/miniconda3/envs/diffusion/include/python3.8 -D_GLIBCXX_USE_CXX11_ABI=0 -D__CUDA_NO_HALF_OPERATORS -D__CUDA_NO_HALF_CONVERSIONS_ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr -gencode=arch=compute_80,code=compute_80 -gencode=arch=compute_80,code=sm_80 --compiler-options '-fPIC' -std=c++14 -c /data/home/zhourui/code/sgmse-main-rev/sgmse/backbones/ncsnpp_utils/op/fused_bias_act_kernel.cu -o fused_bias_act_kernel.cuda.o
/usr/include/c++/11/type_traits(1406): error: type name is not allowed