DeepSpeed [BUG] whl does not get created following the instructions on Windows 11

build_win.bat Administrative permissions required. Detecting permissions... Success: Administrative permissions confirmed. NOTE: Redirects are currently not supported in Windows or MacOs. DS_BUILD_OPS=1 test.c LINK : fatal error LNK1181: cannot open input file 'aio.lib' [WARNING] async_io requires the dev libaio .so object and headers but these were not found. [WARNING] If libaio is already installed (perhaps from source), try setting the CFLAGS and LDFLAGS environment variables to where it can be found. [WARNING] please install triton==1.0.0 if you want to use sparse attention Install Ops={'async_io': False, 'cpu_adagrad': 1, 'cpu_adam': 1, 'fused_adam': 1, 'fused_lamb': 1, 'quantizer': 1, 'random_ltd': 1, 'sparse_attn': False, 'spatial_inference': 1, 'transformer': 1, 'stochastic_transformer': 1, 'transformer_inference': 1, 'utils': 1} version=0.9.0+f662bfcd, git_hash=f662bfcd, git_branch=master install_requires=['hjson', 'ninja', 'numpy', 'packaging>=20.0', 'psutil', 'py-cpuinfo', 'pydantic', 'torch', 'tqdm'] compatible_ops={'async_io': False, 'cpu_adagrad': True, 'cpu_adam': True, 'fused_adam': True, 'fused_lamb': True, 'quantizer': True, 'random_ltd': True, 'sparse_attn': False, 'spatial_inference': True, 'transformer': True, 'stochastic_transformer': True, 'transformer_inference': True, 'utils': True} ext_modules=[<setuptools.extension.Extension('deepspeed.ops.adagrad.cpu_adagrad_op') at 0x2c2fb8d69a0>, <setuptools.extension.Extension('deepspeed.ops.adam.cpu_adam_op') at 0x2c2dcd708b0>, <setuptools.extension.Extension('deepspeed.ops.adam.fused_adam_op') at 0x2c2fe9d76a0>, <setuptools.extension.Extension('deepspeed.ops.lamb.fused_lamb_op') at 0x2c2fe9d74f0>, <setuptools.extension.Extension('deepspeed.ops.quantizer.quantizer_op') at 0x2c2fe9d7490>, <setuptools.extension.Extension('deepspeed.ops.random_ltd_op') at 0x2c2fe9d7760>, <setuptools.extension.Extension('deepspeed.ops.spatial.spatial_inference_op') at 0x2c2fe9d7820>, <setuptools.extension.Extension('deepspeed.ops.transformer.transformer_op') at 0x2c2fe9d7520>, <setuptools.extension.Extension('deepspeed.ops.transformer.stochastic_transformer_op') at 0x2c2fe9d7910>, <setuptools.extension.Extension('deepspeed.ops.transformer.inference.transformer_inference_op') at 0x2c2fe9d7880>, <setuptools.extension.Extension('deepspeed.ops.utils_op') at 0x2c2fe9d7670>] running bdist_wheel running build running build_py copying deepspeed\git_version_info_installed.py -> build\lib.win-amd64-cpython-38\deepspeed running egg_info writing deepspeed.egg-info\PKG-INFO writing dependency_links to deepspeed.egg-info\dependency_links.txt writing entry points to deepspeed.egg-info\entry_points.txt writing requirements to deepspeed.egg-info\requires.txt writing top-level names to deepspeed.egg-info\top_level.txt reading manifest file 'deepspeed.egg-info\SOURCES.txt' reading manifest template 'MANIFEST_win.in' warning: no previously-included files matching '.cpp' found under directory 'deepspeed\ops\csrc' warning: no previously-included files matching '.h' found under directory 'deepspeed\ops\csrc' warning: no previously-included files matching '.cu' found under directory 'deepspeed\ops\csrc' warning: no previously-included files matching '.cuh' found under directory 'deepspeed\ops\csrc' warning: no previously-included files matching '*.cc' found under directory 'deepspeed\ops\csrc' no previously-included directories found matching 'op_builder' no previously-included directories found matching 'accelerator' adding license file 'LICENSE' writing manifest file 'deepspeed.egg-info\SOURCES.txt' running build_ext building 'deepspeed.ops.quantizer.quantizer_op' extension "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.6\bin\nvcc" -c csrc/quantization/dequantize.cu -o build\temp.win-amd64-cpython-38\Release\csrc/quantization/dequantize.obj -Icsrc/includes -IC:\Users\amolamb\Anaconda3\envs\asuka\lib\site-packages\torch\include -IC:\Users\amolamb\Anaconda3\envs\asuka\lib\site-packages\torch\include\torch\csrc\api\include -IC:\Users\amolamb\Anaconda3\envs\asuka\lib\site-packages\torch\include\TH -IC:\Users\amolamb\Anaconda3\envs\asuka\lib\site-packages\torch\include\THC "-IC:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.6\include" -IC:\Users\amolamb\Anaconda3\envs\asuka\include -IC:\Users\amolamb\Anaconda3\envs\asuka\Include "-Ic:\Program Files (x86)\Microsoft Visual Studio\2019\BuildTools\VC\Tools\MSVC\14.29.30133\include" "-IC:\Program Files (x86)\Windows Kits\NETFXSDK\4.8\include\um" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.22000.0\ucrt" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.22000.0\shared" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.22000.0\um" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.22000.0\winrt" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.22000.0\cppwinrt" -Xcudafe --diag_suppress=dll_interface_conflict_dllexport_assumed -Xcudafe --

. . . . . . . . . . . . . . . C:\Users\amolamb\Anaconda3\envs\asuka\lib\site-packages\torch\include\pybind11\detail/common.h(108): warning C4005: 'HAVE_SNPRINTF': macro redefinition C:\Users\amolamb\Anaconda3\envs\asuka\include\pyerrors.h(315): note: see previous definition of 'HAVE_SNPRINTF' "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.6\bin\nvcc" -c csrc/quantization/quantize.cu -o build\temp.win-amd64-cpython-38\Release\csrc/quantization/quantize.obj -Icsrc/includes -IC:\Users\amolamb\Anaconda3\envs\asuka\lib\site-packages\torch\include -IC:\Users\amolamb\Anaconda3\envs\asuka\lib\site-packages\torch\include\torch\csrc\api\include -IC:\Users\amolamb\Anaconda3\envs\asuka\lib\site-packages\torch\include\TH -IC:\Users\amolamb\Anaconda3\envs\asuka\lib\site-packages\torch\include\THC "-IC:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.6\include" -IC:\Users\amolamb\Anaconda3\envs\asuka\include -IC:\Users\amolamb\Anaconda3\envs\asuka\Include "-Ic:\Program Files (x86)\Microsoft Visual Studio\2019\BuildTools\VC\Tools\MSVC\14.29.30133\include" "-IC:\Program Files (x86)\Windows Kits\NETFXSDK\4.8\include\um" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.22000.0\ucrt" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.22000.0\shared" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.22000.0\um" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.22000.0\winrt" "-IC:\Program Files (x86)\Windows Kits\10\include\10.0.22000.0\cppwinrt" -Xcudafe --diag_suppress=dll_interface_conflict_dllexport_assumed -Xcudafe --diag_suppress=dll_interface_conflict_none_assumed -Xcudafe --diag_suppress=field_without_dll_interface -Xcudafe --diag_suppress=base_class_has_different_dll_interface -Xcompiler /EHsc -Xcompiler /wd4190 -Xcompiler /wd4018 -Xcompiler /wd4275 -Xcompiler /wd4267 -Xcompiler /wd4244 -Xcompiler /wd4251 -Xcompiler /wd4819 -Xcompiler /MD -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_BFLOAT16_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__ --expt-relaxed-constexpr -O3 -allow-unsupported-compiler --use_fast_math -std=c++17 -U__CUDA_NO_HALF_OPERATORS__ -U__CUDA_NO_HALF_CONVERSIONS__ -U__CUDA_NO_HALF2_OPERATORS__ -gencode=arch=compute_60,code=sm_60 -gencode=arch=compute_61,code=sm_61 -gencode=arch=compute_70,code=sm_70 -gencode=arch=compute_80,code=sm_80 -gencode=arch=compute_86,code=sm_86 -DTORCH_API_INCLUDE_EXTENSION_H -DTORCH_EXTENSION_NAME=quantizer_op -D_GLIBCXX_USE_CXX11_ABI=0 --use-local-env quantize.cu cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF_OPERATORS__' with '/U__CUDA_NO_HALF_OPERATORS__' cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF_CONVERSIONS__' with '/U__CUDA_NO_HALF_CONVERSIONS__' cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF2_OPERATORS__' with '/U__CUDA_NO_HALF2_OPERATORS__' quantize.cu cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF_OPERATORS__' with '/U__CUDA_NO_HALF_OPERATORS__' cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF_CONVERSIONS__' with '/U__CUDA_NO_HALF_CONVERSIONS__' cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF2_OPERATORS__' with '/U__CUDA_NO_HALF2_OPERATORS__' quantize.cu cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF_OPERATORS__' with '/U__CUDA_NO_HALF_OPERATORS__' cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF_CONVERSIONS__' with '/U__CUDA_NO_HALF_CONVERSIONS__' cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF2_OPERATORS__' with '/U__CUDA_NO_HALF2_OPERATORS__' quantize.cu cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF_OPERATORS__' with '/U__CUDA_NO_HALF_OPERATORS__' cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF_CONVERSIONS__' with '/U__CUDA_NO_HALF_CONVERSIONS__' cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF2_OPERATORS__' with '/U__CUDA_NO_HALF2_OPERATORS__' quantize.cu cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF_OPERATORS__' with '/U__CUDA_NO_HALF_OPERATORS__' cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF_CONVERSIONS__' with '/U__CUDA_NO_HALF_CONVERSIONS__' cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF2_OPERATORS__' with '/U__CUDA_NO_HALF2_OPERATORS__' quantize.cu cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF_OPERATORS__' with '/U__CUDA_NO_HALF_OPERATORS__' cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF_CONVERSIONS__' with '/U__CUDA_NO_HALF_CONVERSIONS__' cl : Command line warning D9025 : overriding '/D__CUDA_NO_HALF2_OPERATORS__' with '/U__CUDA_NO_HALF2_OPERATORS__' quantize.cu csrc/includes\memory_access_utils.h(82): error: identifier "uint32_t" is undefined

csrc/includes\memory_access_utils.h(359): error: identifier "int32_t" is undefined

csrc/includes\memory_access_utils.h(359): error: identifier "data" is undefined

csrc/includes\memory_access_utils.h(359): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(363): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(371): error: identifier "int32_t" is undefined

csrc/includes\memory_access_utils.h(371): error: identifier "data" is undefined

csrc/includes\memory_access_utils.h(371): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(383): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(395): error: identifier "int32_t" is undefined

csrc/includes\memory_access_utils.h(395): error: identifier "data" is undefined

csrc/includes\memory_access_utils.h(395): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(399): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(409): error: identifier "int32_t" is undefined

csrc/includes\memory_access_utils.h(409): error: identifier "data" is undefined

csrc/includes\memory_access_utils.h(409): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(421): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(434): error: identifier "int32_t" is undefined

csrc/includes\memory_access_utils.h(434): error: identifier "data" is undefined

csrc/includes\memory_access_utils.h(434): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(438): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(448): error: identifier "int32_t" is undefined

csrc/includes\memory_access_utils.h(448): error: identifier "data" is undefined

csrc/includes\memory_access_utils.h(448): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(460): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(472): error: identifier "int16_t" is undefined

csrc/includes\memory_access_utils.h(472): error: identifier "data" is undefined

csrc/includes\memory_access_utils.h(472): error: "int16_t" is not a type name

csrc/includes\memory_access_utils.h(476): error: "int16_t" is not a type name

csrc/includes\memory_access_utils.h(484): error: identifier "int16_t" is undefined

csrc/includes\memory_access_utils.h(484): error: identifier "data" is undefined

csrc/includes\memory_access_utils.h(484): error: "int16_t" is not a type name

csrc/includes\memory_access_utils.h(496): error: "int16_t" is not a type name

csrc/includes\memory_access_utils.h(508): error: identifier "int16_t" is undefined

csrc/includes\memory_access_utils.h(508): error: identifier "data" is undefined

csrc/includes\memory_access_utils.h(508): error: "int16_t" is not a type name

csrc/includes\memory_access_utils.h(512): error: "int16_t" is not a type name

csrc/includes\memory_access_utils.h(522): error: identifier "int16_t" is undefined

csrc/includes\memory_access_utils.h(522): error: identifier "data" is undefined

csrc/includes\memory_access_utils.h(522): error: "int16_t" is not a type name

csrc/includes\memory_access_utils.h(534): error: "int16_t" is not a type name

csrc/includes\memory_access_utils.h(547): error: identifier "int16_t" is undefined

csrc/includes\memory_access_utils.h(547): error: identifier "data" is undefined

csrc/includes\memory_access_utils.h(547): error: "int16_t" is not a type name

csrc/includes\memory_access_utils.h(551): error: "int16_t" is not a type name

csrc/includes\memory_access_utils.h(561): error: identifier "int16_t" is undefined

csrc/includes\memory_access_utils.h(561): error: identifier "data" is undefined

csrc/includes\memory_access_utils.h(561): error: "int16_t" is not a type name

csrc/includes\memory_access_utils.h(573): error: "int16_t" is not a type name

csrc/includes\memory_access_utils.h(703): error: identifier "int32_t" is undefined

csrc/includes\memory_access_utils.h(703): error: identifier "data" is undefined

csrc/includes\memory_access_utils.h(703): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(709): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(717): error: identifier "int32_t" is undefined

csrc/includes\memory_access_utils.h(717): error: identifier "data" is undefined

csrc/includes\memory_access_utils.h(717): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(731): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(836): error: identifier "int32_t" is undefined

csrc/includes\memory_access_utils.h(840): error: identifier "int32_t" is undefined

csrc/includes\memory_access_utils.h(840): error: identifier "dst_cast" is undefined

csrc/includes\memory_access_utils.h(840): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(849): error: identifier "int32_t" is undefined

csrc/includes\memory_access_utils.h(853): error: identifier "int32_t" is undefined

csrc/includes\memory_access_utils.h(853): error: identifier "dst_cast" is undefined

csrc/includes\memory_access_utils.h(853): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(862): error: identifier "int32_t" is undefined

csrc/includes\memory_access_utils.h(866): error: identifier "int32_t" is undefined

csrc/includes\memory_access_utils.h(866): error: identifier "dst_cast" is undefined

csrc/includes\memory_access_utils.h(866): error: "int32_t" is not a type name

csrc/includes\memory_access_utils.h(908): error: identifier "int32_t" is undefined

csrc/includes\memory_access_utils.h(914): error: identifier "int32_t" is undefined

csrc/includes\memory_access_utils.h(914): error: identifier "dst_cast" is undefined

csrc/includes\memory_access_utils.h(914): error: "int32_t" is not a type name

91 errors detected in the compilation of "csrc/quantization/quantize.cu". error: command 'C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.6\bin\nvcc.exe' failed with exit code 2

Apr 12 '23 06:04 acube3

Hi @acube3 - it looks like one of your errors if from trying to install aio, which is not supported on Windows. Can you share the commands you're running to get to this point, and the environment you have?

Apr 18 '23 23:04 loadams

i have same problem with you @acube3 , did you reslove ?

Apr 20 '23 11:04 abc20220327

@abc20220327 and @acube3, can you please confirm that you have VS2019 C++ x64/x86 build tools, cuda, and torch installed? And the versions? Then once you've done that, can you run the following to confirm they're installed and found by your python instance?

python -c "import torch; print('torch:', torch.__version__, torch)"
python -c "import torch; print('CUDA available:', torch.cuda.is_available())"

I was just able to follow the steps listed on a Windows 11 machine and get it to build.

Apr 20 '23 22:04 loadams

i have same problem , this is my torch `PS F:\DeepSpeed-0.9.1> python -c "import torch; print('torch:', torch.version, torch)" torch: 1.13.0+cu116 <module 'torch' from 'F:\ProgramData\Anaconda3\lib\site-packages\torch\init.py'>

PS F:\DeepSpeed-0.9.1> python -c "import torch; print('CUDA available:', torch.cuda.is_available())" CUDA available: True`

Apr 24 '23 02:04 xfuzq707

@xfuzq707 - thanks, those should be fine. Can you confirm the commands you are running, and the full error message?

Apr 24 '23 13:04 loadams

me too

/** (base) PS E:\deepspeed> python -c "import torch; print('torch:', torch.version, torch)" torch: 2.0.0+cu117 <module 'torch' from 'E:\Anaconda3\lib\site-packages\torch\init.py'> (base) PS E:\deepspeed> python -c "import torch; print('CUDA available:', torch.cuda.is_available())" CUDA available: True */

all is ready,but i have same problem. cannot open input file 'aio.lib'

Apr 25 '23 14:04 shaolongcai

@shaolongcai - aio isn't supported on Windows. If using the build_win.bat script, this is explicitly set - how are you building on Windows? If you want/need aio functionality, you'd be better using WSL on your Windows machine if you can.

Apr 25 '23 15:04 loadams

@xfuzq707 - thanks, those should be fine. Can you confirm the commands you are running, and the full error message?

Yes, there are all message:

`

PS F:\DeepSpeed-0.9.1> .\build_win.bat Administrative permissions required. Detecting permissions... Success: Administrative permissions confirmed. DS_BUILD_OPS=1 test.c LINK : fatal error LNK1181: 无法打开输入文件“aio.lib” [WARNING] async_io requires the dev libaio .so object and headers but these were not found. [WARNING] If libaio is already installed (perhaps from source), try setting the CFLAGS and LDFLAGS environment variables to where it can be found. 系统找不到指定的文件。 [WARNING] cpu_adagrad requires the 'lscpu' command, but it does not exist! [WARNING] cpu_adagrad attempted to query 'lscpu' after failing to use py-cpuinfo to detect the CPU architecture. 'lscpu' does not appear to exist on your system, will fall back to use -march=native and non-vectorized execution. 系统找不到指定的文件。 [WARNING] cpu_adagrad requires the 'lscpu' command, but it does not exist! [WARNING] cpu_adagrad attempted to query 'lscpu' after failing to use py-cpuinfo to detect the CPU architecture. 'lscpu' does not appear to exist on your system, will fall back to use -march=native and non-vectorized execution. 系统找不到指定的文件。 [WARNING] cpu_adam requires the 'lscpu' command, but it does not exist! [WARNING] cpu_adam attempted to query 'lscpu' after failing to use py-cpuinfo to detect the CPU architecture. 'lscpu' does not appear to exist on your system, will fall back to use -march=native and non-vectorized execution. 系统找不到指定的文件。 [WARNING] cpu_adam requires the 'lscpu' command, but it does not exist! [WARNING] cpu_adam attempted to query 'lscpu' after failing to use py-cpuinfo to detect the CPU architecture. 'lscpu' does not appear to exist on your system, will fall back to use -march=native and non-vectorized execution. [WARNING] please install triton==1.0.0 if you want to use sparse attention Install Ops={'async_io': False, 'cpu_adagrad': 1, 'cpu_adam': 1, 'fused_adam': 1, 'fused_lamb': 1, 'quantizer': 1, 'random_ltd': 1, 'sparse_attn': False, 'spatial_inference': 1, 'transformer': 1, 'stochastic_transformer': 1, 'transformer_inference': 1, 'utils': 1} fatal: not a git repository (or any of the parent directories): .git version=0.9.1+unknown, git_hash=unknown, git_branch=unknown install_requires=['hjson', 'ninja', 'numpy', 'packaging>=20.0', 'psutil', 'py-cpuinfo', 'pydantic<2.0.0', 'torch', 'tqdm']

........

quantize.cu cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF_OPERATORS__”(用“/U__CUDA_NO_HALF_OPERATORS__”) cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF_CONVERSIONS__”(用“/U__CUDA_NO_HALF_CONVERSIONS__”) cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF2_OPERATORS__”(用“/U__CUDA_NO_HALF2_OPERATORS__”) quantize.cu cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF_OPERATORS__”(用“/U__CUDA_NO_HALF_OPERATORS__”) cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF_CONVERSIONS__”(用“/U__CUDA_NO_HALF_CONVERSIONS__”) cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF2_OPERATORS__”(用“/U__CUDA_NO_HALF2_OPERATORS__”) quantize.cu cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF_OPERATORS__”(用“/U__CUDA_NO_HALF_OPERATORS__”) cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF_CONVERSIONS__”(用“/U__CUDA_NO_HALF_CONVERSIONS__”) cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF2_OPERATORS__”(用“/U__CUDA_NO_HALF2_OPERATORS__”) quantize.cu cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF_OPERATORS__”(用“/U__CUDA_NO_HALF_OPERATORS__”) cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF_CONVERSIONS__”(用“/U__CUDA_NO_HALF_CONVERSIONS__”) cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF2_OPERATORS__”(用“/U__CUDA_NO_HALF2_OPERATORS__”) quantize.cu cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF_OPERATORS__”(用“/U__CUDA_NO_HALF_OPERATORS__”) cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF_CONVERSIONS__”(用“/U__CUDA_NO_HALF_CONVERSIONS__”) cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF2_OPERATORS__”(用“/U__CUDA_NO_HALF2_OPERATORS__”) quantize.cu cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF_OPERATORS__”(用“/U__CUDA_NO_HALF_OPERATORS__”) cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF_CONVERSIONS__”(用“/U__CUDA_NO_HALF_CONVERSIONS__”) cl: 命令行 warning D9025 :正在重写“/D__CUDA_NO_HALF2_OPERATORS__”(用“/U__CUDA_NO_HALF2_OPERATORS__”) quantize.cu csrc/includes\memory_access_utils.h(82): error: identifier "uint32_t" is undefined