Installation for amazon DL1 EC2 instances (Gaudi processors, Habana labs)

[dhall1995]

😵 Describe the installation problem

I recently gained access to an amazon EC2 DL1 instance which use Gaudi accelerators rather than nvidia. I was wondering whether it is even possible to install pytorch geometric with this accelerator or if this will just be impossible with any current torch geometric version

Environment

• PyG version: cpu
• PyTorch version: 1.12.0
• OS: DISTRIB_ID=Ubuntu DISTRIB_RELEASE=20.04 DISTRIB_CODENAME=focal DISTRIB_DESCRIPTION="Ubuntu 20.04.5 LTS" NAME="Ubuntu" VERSION="20.04.5 LTS (Focal Fossa)" ID=ubuntu ID_LIKE=debian PRETTY_NAME="Ubuntu 20.04.5 LTS" VERSION_ID="20.04" HOME_URL="https://www.ubuntu.com/" SUPPORT_URL="https://help.ubuntu.com/" BUG_REPORT_URL="https://bugs.launchpad.net/ubuntu/" PRIVACY_POLICY_URL="https://www.ubuntu.com/legal/terms-and-policies/privacy-policy" VERSION_CODENAME=focal UBUNTU_CODENAME=focal
• Python version: 3.8.10
• CUDA/cuDNN version: na
• How you installed PyTorch and PyG (conda, pip, source): pip
• Any other relevant information (e.g., version of torch-scatter): I ran - pip install pyg-lib torch-scatter torch-sparse -f https://data.pyg.org/whl/torch-1.12.0+cpu.html

[dhall1995]

Oh sorry I should add - the error when installing pytorch geometric is:

>>> import torch_geometric as pyg
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/ubuntu/.local/lib/python3.8/site-packages/torch_geometric/__init__.py", line 4, in <module>
    import torch_geometric.data
  File "/home/ubuntu/.local/lib/python3.8/site-packages/torch_geometric/data/__init__.py", line 1, in <module>
    from .data import Data
  File "/home/ubuntu/.local/lib/python3.8/site-packages/torch_geometric/data/data.py", line 20, in <module>
    from torch_sparse import SparseTensor
  File "/home/ubuntu/.local/lib/python3.8/site-packages/torch_sparse/__init__.py", line 18, in <module>
    torch.ops.load_library(spec.origin)
  File "/usr/local/lib/python3.8/dist-packages/torch/_ops.py", line 255, in load_library
    ctypes.CDLL(path)
  File "/usr/lib/python3.8/ctypes/__init__.py", line 373, in __init__
    self._handle = _dlopen(self._name, mode)
OSError: /home/ubuntu/.local/lib/python3.8/site-packages/torch_sparse/_version_cpu.so: undefined symbol: _ZN5torch3jit17parseSchemaOrNameERKSs

[rusty1s]

We just dropped any requirements of dependencies of torch-scatter and torch-sparse and PyG can now run now natively using PyTorch. Can you try to install PyG from master without any of these extensions?