Cannot build wheel

[LankyPoet]

Hi,
I am running Windows 11, Python 3.11.9, and comfyui in a venv environment.
I tried installing the latest llama-cpp-python for Cuda 1.24 in the below manner and received a string of errors. Can anyone assist please?

(venv) D:\ComfyUI>pip install llama-cpp-python --verbose --extra-index-url https://abetlen.github.io/llama-cpp-python/wh
l/cu124
Using pip 24.0 from D:\ComfyUI\venv\Lib\site-packages\pip (python 3.11)
Looking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com, https://abetlen.github.io/llama-cpp-python/whl/cu124
Collecting llama-cpp-python
  Downloading llama_cpp_python-0.2.78.tar.gz (50.2 MB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 50.2/50.2 MB 32.7 MB/s eta 0:00:00
  Running command pip subprocess to install build dependencies
  Looking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com, https://pypi.ngc.nvidia.com, https://abetlen.github.io/llama-cpp-python/whl/cu124
  Collecting scikit-build-core>=0.9.2 (from scikit-build-core[pyproject]>=0.9.2)
    Downloading scikit_build_core-0.9.6-py3-none-any.whl.metadata (19 kB)
  Collecting packaging>=21.3 (from scikit-build-core>=0.9.2->scikit-build-core[pyproject]>=0.9.2)
    Downloading packaging-24.1-py3-none-any.whl.metadata (3.2 kB)
  Collecting pathspec>=0.10.1 (from scikit-build-core>=0.9.2->scikit-build-core[pyproject]>=0.9.2)
    Downloading pathspec-0.12.1-py3-none-any.whl.metadata (21 kB)
  Downloading scikit_build_core-0.9.6-py3-none-any.whl (152 kB)
     ---------------------------------------- 152.3/152.3 kB 4.6 MB/s eta 0:00:00
  Downloading packaging-24.1-py3-none-any.whl (53 kB)
     ---------------------------------------- 54.0/54.0 kB ? eta 0:00:00
  Downloading pathspec-0.12.1-py3-none-any.whl (31 kB)
  Installing collected packages: pathspec, packaging, scikit-build-core
  Successfully installed packaging-24.1 pathspec-0.12.1 scikit-build-core-0.9.6
  Installing build dependencies ... done
  Running command Getting requirements to build wheel
  Could not determine CMake version via --version, got '' 'Traceback (most recent call last):\n  File "<frozen runpy>", line 198, in _run_module_as_main\n  File "<frozen runpy>", line 88, in _run_code\n  File "D:\\ComfyUI\\venv\\Scripts\\cmake.EXE\\__main__.py", line 4, in <module>\nModuleNotFoundError: No module named \'cmake\'\n'
  Getting requirements to build wheel ... done
  Running command pip subprocess to install backend dependencies
  Looking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com, https://pypi.ngc.nvidia.com, https://abetlen.github.io/llama-cpp-python/whl/cu124
  Collecting cmake>=3.21
    Downloading cmake-3.29.5.1-py3-none-win_amd64.whl.metadata (6.1 kB)
  Downloading cmake-3.29.5.1-py3-none-win_amd64.whl (36.2 MB)
     ---------------------------------------- 36.2/36.2 MB 65.2 MB/s eta 0:00:00
  Installing collected packages: cmake
  Successfully installed cmake-3.29.5.1
  Installing backend dependencies ... done
  Running command Preparing metadata (pyproject.toml)
  *** scikit-build-core 0.9.6 using CMake 3.29.5 (metadata_wheel)
  Preparing metadata (pyproject.toml) ... done
Requirement already satisfied: typing-extensions>=4.5.0 in d:\comfyui\venv\lib\site-packages (from llama-cpp-python) (4.12.2)
Requirement already satisfied: numpy>=1.20.0 in d:\comfyui\venv\lib\site-packages (from llama-cpp-python) (1.26.4)
Requirement already satisfied: diskcache>=5.6.1 in d:\comfyui\venv\lib\site-packages (from llama-cpp-python) (5.6.3)
Requirement already satisfied: jinja2>=2.11.3 in d:\comfyui\venv\lib\site-packages (from llama-cpp-python) (3.1.4)
Requirement already satisfied: MarkupSafe>=2.0 in d:\comfyui\venv\lib\site-packages (from jinja2>=2.11.3->llama-cpp-python) (2.1.5)
Building wheels for collected packages: llama-cpp-python
  Running command Building wheel for llama-cpp-python (pyproject.toml)
  *** scikit-build-core 0.9.6 using CMake 3.29.5 (wheel)
  *** Configuring CMake...
  2024-06-17 12:05:30,337 - scikit_build_core - WARNING - Can't find a Python library, got libdir=None, ldlibrary=None, multiarch=None, masd=None
  loading initial cache file C:\Users\DEFAUL~1.LIV\AppData\Local\Temp\tmpy2u4tmwa\build\CMakeInit.txt
  -- Building for: Visual Studio 17 2022
  -- Selecting Windows SDK version 10.0.22621.0 to target Windows 10.0.22631.
  -- The C compiler identification is MSVC 19.39.33520.0
  -- The CXX compiler identification is MSVC 19.39.33520.0
  -- Detecting C compiler ABI info
  -- Detecting C compiler ABI info - done
  -- Check for working C compiler: C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.39.33519/bin/Hostx64/x64/cl.exe - skipped
  -- Detecting C compile features
  -- Detecting C compile features - done
  -- Detecting CXX compiler ABI info
  -- Detecting CXX compiler ABI info - done
  -- Check for working CXX compiler: C:/Program Files (x86)/Microsoft Visual Studio/2022/BuildTools/VC/Tools/MSVC/14.39.33519/bin/Hostx64/x64/cl.exe - skipped
  -- Detecting CXX compile features
  -- Detecting CXX compile features - done
  -- Found Git: C:/Program Files/Git/cmd/git.exe (found version "2.45.2.windows.1")
  -- Performing Test CMAKE_HAVE_LIBC_PTHREAD
  -- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Failed
  -- Looking for pthread_create in pthreads
  -- Looking for pthread_create in pthreads - not found
  -- Looking for pthread_create in pthread
  -- Looking for pthread_create in pthread - not found
  -- Found Threads: TRUE
  -- Found OpenMP_C: -openmp (found version "2.0")
  -- Found OpenMP_CXX: -openmp (found version "2.0")
  -- Found OpenMP: TRUE (found version "2.0")
  -- OpenMP found
  -- Found CUDAToolkit: C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v12.4/include (found version "12.4.131")
  -- CUDA found
  CMake Error at C:/Users/Default.LivingRoomPC/AppData/Local/Temp/pip-build-env-9sqow3ao/normal/Lib/site-packages/cmake/data/share/cmake-3.29/Modules/CMakeDetermineCompilerId.cmake:563 (message):
    No CUDA toolset found.
  Call Stack (most recent call first):
    C:/Users/Default.LivingRoomPC/AppData/Local/Temp/pip-build-env-9sqow3ao/normal/Lib/site-packages/cmake/data/share/cmake-3.29/Modules/CMakeDetermineCompilerId.cmake:8 (CMAKE_DETERMINE_COMPILER_ID_BUILD)
    C:/Users/Default.LivingRoomPC/AppData/Local/Temp/pip-build-env-9sqow3ao/normal/Lib/site-packages/cmake/data/share/cmake-3.29/Modules/CMakeDetermineCompilerId.cmake:53 (__determine_compiler_id_test)
    C:/Users/Default.LivingRoomPC/AppData/Local/Temp/pip-build-env-9sqow3ao/normal/Lib/site-packages/cmake/data/share/cmake-3.29/Modules/CMakeDetermineCUDACompiler.cmake:131 (CMAKE_DETERMINE_COMPILER_ID)
    vendor/llama.cpp/CMakeLists.txt:411 (enable_language)


  -- Configuring incomplete, errors occurred!

  *** CMake configuration failed
  error: subprocess-exited-with-error

  × Building wheel for llama-cpp-python (pyproject.toml) did not run successfully.
  │ exit code: 1
  ╰─> See above for output.

  note: This error originates from a subprocess, and is likely not a problem with pip.
  full command: 'D:\ComfyUI\venv\Scripts\python.exe' 'D:\ComfyUI\venv\Lib\site-packages\pip\_vendor\pyproject_hooks\_in_process\_in_process.py' build_wheel 'C:\Users\DEFAUL~1.LIV\AppData\Local\Temp\tmpa1tcm9ca'
  cwd: C:\Users\Default.LivingRoomPC\AppData\Local\Temp\pip-install-gn5_hb5u\llama-cpp-python_11e565ea5e874456a27660546fbad291
  Building wheel for llama-cpp-python (pyproject.toml) ... error
  ERROR: Failed building wheel for llama-cpp-python
Failed to build llama-cpp-python
ERROR: Could not build wheels for llama-cpp-python, which is required to install pyproject.toml-based projects

[AspiringPianist]

Yes, same issue. Using a python 3.10.6 virtual environment

[dreambottle]

check this reply

#1535 (comment)

[LankyPoet]

check this reply

#1535 (comment)

Thank you. Not a bad workaround to get going, but I agree with you, I am really hoping we keep seeing updated CUDA builds. New models come out constantly so it's important to stay up on llama.cpp versions.

[javierxio]

This is how i got it to work. Took me a day to figure out.

• Uninstall build tools (standalone version)
• From Visual Studio, install the Desktop Development with C++.
• Reinstall CUDA
• run the pip install

#1352 (comment)