About using your pretrained model on pure point cloud data without color/intensity information

The usage example shows that "coord", "color", "normal" are all required to use your pretrained model.
Can I use it with the input data of pure point cloud (i.e. with only "coord" field)?

Start with the usage example and trace how the pretrained model consumes the coord, color, and normal fields. Determine whether coord-only point clouds are supported; done means the supported input requirements are established and either the example or implementation clearly reflects them.

Written by the indexing model from the issue text.