NVlabs / tiny-cuda-nn

Lightning fast C++/CUDA neural network framework
Other
3.75k stars 454 forks source link

Issue with installing python bindings #376

Open learn2phoenix opened 1 year ago

learn2phoenix commented 1 year ago

When I try to import tinycudann I get the following error:

ImportError: */anaconda3/lib/python3.9/site-packages/tinycudann_bindings/_86_C.cpython-39-x86_64-linux-gnu.so: undefined symbol: _ZNSt15__exception_ptr13exception_ptr9_M_addrefEv

Could someone help, please?

The following is my build setting if it helps

-- The CXX compiler identification is GNU 11.2.0
-- The CUDA compiler identification is NVIDIA 11.7.64
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /opt/local/stow/gcc-11.2.0/bin/g++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Detecting CUDA compiler ABI info
-- Detecting CUDA compiler ABI info - done
-- Check for working CUDA compiler: /opt/common/cuda/cuda-11.7.0/bin/nvcc - skipped
-- Detecting CUDA compile features
-- Detecting CUDA compile features - done
-- Obtained CUDA architectures automatically from installed GPUs
-- Targeting CUDA architectures: 86
-- Module support is disabled.
-- Version: 9.1.1
-- Build type: RelWithDebInfo
jkulhanek commented 7 months ago

This happens with incorrect _GLIBCXX_USE_CXX11_ABI. Is your PyTorch compiled with _GLIBCXX_USE_CXX11_ABI=0 or _GLIBCXX_USE_CXX11_ABI=1?

NoctemNana commented 6 months ago

@learn2phoenix I unfortunatley encountered the same Error and cannot solve it. Did you by any Chance solve it?