Import MessagePassing raise TypeError: metaclass conflict

[deardeer7]

🐛 Describe the bug

Hello, I got a metaclass confit error:

from torch_geometric.nn import MessagePassing
 File "/home/cuda/anaconda3/envs/esm+/lib/python3.8/site-packages/torch_geometric/__init__.py", line 9, in <module>
   import torch_geometric.data
 File "/home/cuda/anaconda3/envs/esm+/lib/python3.8/site-packages/torch_geometric/data/__init__.py", line 7, in <module>
   from .batch import Batch
 File "/home/cuda/anaconda3/envs/esm+/lib/python3.8/site-packages/torch_geometric/data/batch.py", line 12, in <module>
   from torch_geometric.data.dataset import IndexType
 File "/home/cuda/anaconda3/envs/esm+/lib/python3.8/site-packages/torch_geometric/data/dataset.py", line 30, in <module>
   class Dataset(torch.utils.data.Dataset, ABC):
TypeError: metaclass conflict: the metaclass of a derived class must be a (non-strict) subclass of the metaclasses of all its bases

I find same issue here in pytorch issues, but still not clear how to solve, please help check.

Thanks in advance.

Versions

python 3.8.18
[pip3] numpy==1.24.3
[pip3] torch==1.12.1
[pip3] torch-cluster==1.6.3
[pip3] torch_geometric==2.5.0
[pip3] torch-scatter==2.1.2
[pip3] torchaudio==0.12.1
[pip3] torchdrug==0.2.1
[pip3] torchvision==0.13.1
[conda] blas 1.0 mkl defaults
[conda] cudatoolkit 11.6.0 hecad31d_10 conda-forge
[conda] ffmpeg 4.3 hf484d3e_0 pytorch
[conda] mkl 2021.4.0 h06a4308_640 defaults
[conda] mkl-service 2.4.0 py38h95df7f1_0 conda-forge
[conda] mkl_fft 1.3.1 py38h8666266_1 conda-forge
[conda] mkl_random 1.2.2 py38h1abd341_0 conda-forge
[conda] numpy 1.24.3 py38h14f4228_0 defaults
[conda] numpy-base 1.24.3 py38h31eccc5_0 defaults
[conda] pyg 2.5.0 py38_torch_1.12.0_cu116 pyg
[conda] pytorch 1.12.1 py3.8_cuda11.6_cudnn8.3.2_0 pytorch
[conda] pytorch-mutex 1.0 cuda pytorch
[conda] torch-cluster 1.6.3 pypi_0 pypi
[conda] torch-scatter 2.1.2 pypi_0 pypi
[conda] torchaudio 0.12.1 py38_cu116 pytorch
[conda] torchdrug 0.2.1 pypi_0 pypi
[conda] torchvision 0.13.1 py38_cu116 pytorch

[deardeer7]

I delete all ABC class reported in error. It seems worked, but I don't konw why.

class Dataset(torch.utils.data.Dataset, ABC) -> class Dataset(torch.utils.data.Dataset)

[rusty1s]

Thanks. I don't have strong preference to keep the ABC, so I made #8999 to remove it.