Installation error no matter the type of installation

[MaximeBinot]

I can't install the engine no matter the way I try (conda, pip, manual)

I'm currently trying manually as explained in the "System python" part. When trying to execute : python setup.py install I get this error and don't find a way to solve this since I am folowing strictly the guidelines and the requirements

File "setup.py", line 106 assert "=" in arr[0], f"{arr[0]} requires a value." ^ SyntaxError: invalid syntax `` When trying to install with conda I get other compatibility errors not allowing me to install matching versions of cuda, pytorch,...

Is there a more generic way to make sure the installation works ?

Thanks for the help

[UUnforgettable]

https://blog.csdn.net/weixin_44330874/article/details/129678041 I had the same problem，but finally worked，following this。

[vivekmadhavaram]

@UUnforgettable, Even by following the steps in that site, I get error in python setup.py install step and the error is "RuntimeError : Error compiling objects for extension"

[MaximeBinot]

I still couldn't find a way to install the engine. I tried docker but it makes my machine crash for no reason(an Ubuntu20.04 VM but running with more than enough ressources)

I tried again with conda but at this instruction "conda install pytorch=1.9.0 torchvision cudatoolkit=10.2 -c pytorch -c nvidia" I get an error like this with no reasons and can't find a way to fix it

conda install pytorch=1.9.0 torchvision cudatoolkit=10.2 -c pytorch -c nvidia
Collecting package metadata (current_repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Collecting package metadata (repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Solving environment: \ 
Found conflicts! Looking for incompatible packages.

The end of the error message is this but I don't know why I have this much incompatibilities

The following specifications were found to be incompatible with your system:

  - feature:/linux-64::__glibc==2.31=0
  - feature:|@/linux-64::__glibc==2.31=0
  - cudatoolkit=10.2 -> libgcc-ng[version='>=9.3.0'] -> __glibc[version='>=2.17']
  - libffi -> libgcc-ng[version='>=11.2.0'] -> __glibc[version='>=2.17']
  - libgcc-ng -> __glibc[version='>=2.17']
  - libgfortran-ng -> __glibc[version='>=2.17']
  - libopenblas -> libgcc-ng[version='>=11.2.0'] -> __glibc[version='>=2.17']
  - libstdcxx-ng -> __glibc[version='>=2.17']
  - ncurses -> libgcc-ng[version='>=11.2.0'] -> __glibc[version='>=2.17']
  - openssl -> libgcc-ng[version='>=7.5.0'] -> __glibc[version='>=2.17']
  - python=3.8 -> libgcc-ng[version='>=11.2.0'] -> __glibc[version='>=2.17']
  - pytorch=1.9.0 -> cudatoolkit[version='>=11.1,<11.2'] -> __glibc[version='>=2.17,<3.0.a0']
  - readline -> libgcc-ng[version='>=11.2.0'] -> __glibc[version='>=2.17']
  - sqlite -> libgcc-ng[version='>=11.2.0'] -> __glibc[version='>=2.17']
  - tk -> libgcc-ng[version='>=7.5.0'] -> __glibc[version='>=2.17']
  - torchvision -> __glibc[version='>=2.17,<3.0.a0']
  - torchvision -> libgcc-ng[version='>=11.2.0'] -> __glibc[version='>=2.17']
  - xz -> libgcc-ng[version='>=11.2.0'] -> __glibc[version='>=2.17']

Your installed version is: 2.31hen following exactly the installation.

[UUnforgettable]

I still couldn't find a way to install the engine. I tried docker but it makes my machine crash for no reason(an Ubuntu20.04 VM but running with more than enough ressources)

I tried again with conda but at this instruction "conda install pytorch=1.9.0 torchvision cudatoolkit=10.2 -c pytorch -c nvidia" I get an error like this with no reasons and can't find a way to fix it

conda install pytorch=1.9.0 torchvision cudatoolkit=10.2 -c pytorch -c nvidia
Collecting package metadata (current_repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Collecting package metadata (repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Solving environment: \ 
Found conflicts! Looking for incompatible packages.

The end of the error message is this but I don't know why I have this much incompatibilities

The following specifications were found to be incompatible with your system:

  - feature:/linux-64::__glibc==2.31=0
  - feature:|@/linux-64::__glibc==2.31=0
  - cudatoolkit=10.2 -> libgcc-ng[version='>=9.3.0'] -> __glibc[version='>=2.17']
  - libffi -> libgcc-ng[version='>=11.2.0'] -> __glibc[version='>=2.17']
  - libgcc-ng -> __glibc[version='>=2.17']
  - libgfortran-ng -> __glibc[version='>=2.17']
  - libopenblas -> libgcc-ng[version='>=11.2.0'] -> __glibc[version='>=2.17']
  - libstdcxx-ng -> __glibc[version='>=2.17']
  - ncurses -> libgcc-ng[version='>=11.2.0'] -> __glibc[version='>=2.17']
  - openssl -> libgcc-ng[version='>=7.5.0'] -> __glibc[version='>=2.17']
  - python=3.8 -> libgcc-ng[version='>=11.2.0'] -> __glibc[version='>=2.17']
  - pytorch=1.9.0 -> cudatoolkit[version='>=11.1,<11.2'] -> __glibc[version='>=2.17,<3.0.a0']
  - readline -> libgcc-ng[version='>=11.2.0'] -> __glibc[version='>=2.17']
  - sqlite -> libgcc-ng[version='>=11.2.0'] -> __glibc[version='>=2.17']
  - tk -> libgcc-ng[version='>=7.5.0'] -> __glibc[version='>=2.17']
  - torchvision -> __glibc[version='>=2.17,<3.0.a0']
  - torchvision -> libgcc-ng[version='>=11.2.0'] -> __glibc[version='>=2.17']
  - xz -> libgcc-ng[version='>=11.2.0'] -> __glibc[version='>=2.17']

Your installed version is: 2.31hen following exactly the installation.

conda create -n mink python=3.8 conda activate mink pip install torch==1.12.1+cu116 torchvision==0.13.1+cu116 torchaudio==0.12.1 --extra-index-url https://download.pytorch.org/whl/cu116 conda install openblas-devel -c anaconda export CUDA_HOME=/usr/local/cuda-11.6

make sure gcc --version >=7.5.0

g++ --version >=7.5.0

git clone https://github.com/NVIDIA/MinkowskiEngine.git cd MinkowskiEngine python setup.py install --blas_include_dirs=${CONDA_PREFIX}/include --blas=openblas

test

python

import MinkowskiEngine as ME print(ME.version) 0.5.4

try this

[UUnforgettable]

@UUnforgettable, Even by following the steps in that site, I get error in python setup.py install step and the error is "RuntimeError : Error compiling objects for extension" Have you confirmed the version of gcc and g++？

[vivekmadhavaram]

@UUnforgettable, Even by following the steps in that site, I get error in python setup.py install step and the error is "RuntimeError : Error compiling objects for extension" Have you confirmed the version of gcc and g++？

Yeah, both are of version 7.5

and Cuda is of version 11.6