SI_LULC Pipeline Grouth Truth Data

I have seen that in the example you use

land_use_10class_reference_slovenia_partial.gpkg
or land_use_10class_reference_slovenia.gpkg for the full version.

for the LULC_REFERENCE as a ground truth data, so I have a couple of questions:

1. How do you find the ground truth data from other countries, did you label it by yourself, or you find a reference from somewhere and convert it into gpkg format ?
2. My other question is can we use scene classification layer produced by Sen2Cor, ‘SCL’ for the ground truth data instead ?