What requirements do I need to use Torch.compile?

torch.compile
1

Hi, everyone.

I’m using Python 3.13.9.
With PyTorch 2.9 and CUDA 13.0, my GPU is an RTX 380 12 GB.
from this link
pip3 install torch torchvision --index-url https://download.pytorch.org/whl/cu130

But I can’t use it; I’m always getting an error.

Exception has occurred: InductorError

CppCompileError: C++ compile error Command: cl /I c:/Users/Emad Younan/AppData/Local/Programs/Python/Python313/Include /I c:/Users/Emad Younan/AppData/Local/Programs/Python/Python313/Lib/site-packages/torch/include /I c:/Users/Emad Younan/AppData/Local/Programs/Python/Python313/Lib/site-packages/torch/include/torch/csrc/api/include /D NOMINMAX /D TORCH_INDUCTOR_CPP_WRAPPER /D STANDALONE_TORCH_HEADER /D C10_USING_CUSTOM_GENERATED_MACROS /O2 /DLL /MD /std:c++20 /wd4819 /wd4251 /wd4244 /wd4267 /wd4275 /wd4018 /wd4190 /wd4624 /wd4067 /wd4068 /EHsc /Zc:__cplusplus /permissive- /openmp /openmp:experimental C:/temp/torch_compile/bi/cbil2ud2wplsgzj6esiu72j2t7zq6phvrdyun5pl56vn2g26y5qg.main.cpp /FeC:/temp/torch_compile/bi/cbil2ud2wplsgzj6esiu72j2t7zq6phvrdyun5pl56vn2g26y5qg.main.pyd /LD /link /LIBPATH:c:/Users/Emad Younan/AppData/Local/Programs/Python/Python313/libs /LIBPATH:c:/Users/Emad Younan/AppData/Local/Programs/Python/Python313/Lib/site-packages/torch/lib torch.lib torch_cpu.lib torch_python.lib sleef.lib c10.lib Output: Microsoft (R) C/C++ Optimizing Compiler Version 19.44.35219 for x64 Copyright (C) Microsoft Corporation. All rights reserved. cl : Command line warning D9025 : overriding ‘/openmp’ with ‘/openmp:experimental’ cbil2ud2wplsgzj6esiu72j2t7zq6phvrdyun5pl56vn2g26y5qg.main.cpp c:/Users/Emad Younan/AppData/Local/Programs/Python/Python313/Lib/site-packages/torch/include\torch/csrc/inductor/cpp_prefix.h(3): fatal error C1083: Cannot open include file: ‘omp.h’: No such file or directory

  • I followed up on those instructions, but still getting the same error
  • MSVC v143 - VS 2022 C++ x64/x86 build tools
  • Windows 11 SDK
  • C++ CMake tools for Windows
  • C++ core features
That’s the script I used.

import os
import torch

# Create a temp directory without spaces

temp_dir = r"C:\temp\torch_compile"

os.makedirs(temp_dir, exist_ok=True)

# Set PyTorch to use this temp directory

os.environ['TORCHINDUCTOR_CACHE_DIR'] = temp_dir

os.environ['TORCH_COMPILE_DEBUG'] = '0'

# Add MSVC to PATH

msvc_path = r"C:\Program Files\Microsoft Visual Studio\2022\Enterprise\VC\Tools\MSVC\14.44.35207\bin\Hostx64\x64"

os.environ["PATH"] = msvc_path + ";" + os.environ["PATH"]

def foo(x, y):

    a = torch.sin(x)

    b = torch.cos(y)

    return a + b

opt_foo = torch.compile(foo)

result = opt_foo(torch.randn(10, 10), torch.randn(10, 10))

print("Success! Result shape:", result.shape)