I am using pytorch-geometric2.0.3, there are a quetion about that 'CGConv' object has no attribute 'decomposed_layers'

[userchba]

Traceback (most recent call last):
File "/home/hbchen/miniconda3/bin/deeph-inference", line 8, in
sys.exit(main())
File "/home/hbchen/miniconda3/lib/python3.9/site-packages/deeph/scripts/inference.py", line 124, in main
predict(input_dir=work_dir, output_dir=work_dir, disable_cuda=disable_cuda, device=device,
File "/home/hbchen/miniconda3/lib/python3.9/site-packages/deeph/inference/pred_ham.py", line 123, in predict
output = kernel.model(batch.x.to(kernel.device), batch.edge_index.to(kernel.device),
File "/home/hbchen/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
return forward_call(*input, **kwargs)
File "<torch_package_0>.deeph/model.py", line 633, in forward
File "/home/hbchen/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
return forward_call(*input, **kwargs)
File "<torch_package_0>.deeph/model.py", line 406, in forward
File "/home/hbchen/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
return forward_call(*input, **kwargs)
File "<torch_package_0>.deeph/model.py", line 145, in forward
File "/home/hbchen/miniconda3/lib/python3.9/site-packages/torch_geometric/nn/conv/message_passing.py", line 263, in propagate
decomposed_layers = 1 if self.explain else self.decomposed_layers
File "/home/hbchen/miniconda3/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1130, in getattr
raise AttributeError("'{}' object has no attribute '{}'".format(
AttributeError: 'CGConv' object has no attribute 'decomposed_layers'

[rusty1s]

Do you load a model that uses GCNConv from disk? Can you upgrade your PyG installation to see if the issue is resolved?