Import error: undefined symbol in executing "Run Pytorch optimizer"

[Danshi-Li]

Hi, I have been working to build this project on my Ubuntu, followed all previous instructions except skipping sudo add-apt-repository ppa:bzindovic/suitesparse-bugfix-1319687. Everything builds smoothly, until when I run the script 4 in Run Pytorch optimizer. It raised the following error:

Traceback (most recent call last):
  File "../src/python/cad_neural_deform2.py", line 10, in <module>
    from layers.graph_loss2_layer import GraphLoss2Layer, Finalize
  File "/home/miyan/danshili/MeshODE/src/python/layers/graph_loss2_layer.py", line 5, in <module>
    import pyDeform
ImportError: /home/miyan/danshili/MeshODE/build/pyDeform.cpython-38-x86_64-linux-gnu.so: undefined symbol: _ZTVN5torch8autograd12AutogradMetaE

In my point of view this is most likely caused by an incompatibility issue between some implementation of this project and my local version of PyTorch with Python=3.8. I have tried the following versions but all failed with the same error: conda install pytorch==1.13.1 torchvision==0.14.1 torchaudio==0.13.1 pytorch-cuda=11.6 -c pytorch -c nvidia 'conda install pytorch==1.10.0 torchvision==0.11.0 torchaudio==0.10.0 cudatoolkit=11.3 -c pytorch -c conda-forge' 'pip install torch==1.10.1+cu111 torchvision==0.11.2+cu111 torchaudio==0.10.1 -f https://download.pytorch.org/whl/cu111/torch_stable.html'

As my GPU hardware requires coda toolkit 11+, using still more previous versions of PyTorch is not viable. Could you give a hint as on which environment settings may this project run successfully? Thanks a lot for your help!

Best, Stanley