Inference result problem

[ldj0o]

Thank you very much for sharing the work. I have built my own dataset and completed the training of part_degmentation. Now I want to use the trained model for inference. I want to input a txt file without a label and output the inference label after inference. However, it seems that the trained model requires four scales of input, which means that the point cloud file I input also needs to have a label. However, the files that need to be inferred often do not have label information and need to infer the label. Can I infer the label by inputting a point cloud file without a label? Thank you again for your work and I am looking forward to your response.