BrokenPipeError initially but after running the cell again in jupyter notebook become AssertionError

[yvonneng996]

Hi,

I have a query regarding the training using the pretrained model on my dataset. Currently, I am using 'upernet_swin_base_patch4_window7_512x512_160k_ade20k.py' as my config file and 'upernet_swin_base_patch4_window7_512x512.pth' and this as the pretrained model.

When I run the config files and the pretrained model right, I am quite curious which swin transformer folders are your using for the backbones. They is one in the Swin-Transformer Semantic Segmentation and the other one is inside this openmmlab: mmsegmentation. For the backbones, which one should I look into?
What does the BrokenPipeError means in details? After I run the cell in jupyter notebook, the error become AssertionError. What exactly does this error mean?
Really need help from your. Thank You