Hello everyone,
I decided to create this issue because I found a solution to the following error: "UnicodeDecodeError: 'ascii' codec can't decode byte 0x8e in position 0: ordinal not in range(128)".
To avoid this error, you just have to look at the documentation https://pytorch.org/docs/master/generated/torch.load.html#torch.load for torch.load and see in the note part that you have to add the argument encoding='latin1'. So, replace in the codes visualize.py and scannetv2_instance.py the torch.load(i) by torch.load(i,encoding='latin1'). Normally, the problem will be solved.