The split file of homologous settings and heterogeneous setting

[flandrewries]

we notice that there are two types of setting in your paper: homologous settings and heterogeneous setting. the homologous settings can be easily obtained by random selection from original train.txt and val.txt. but the heterogeneous setting are not given. we have checked the files of original train.txt and val.txt, and discovered that there are some same scenarios inside them. so how to split them into heterogeneous train_set and val_set? could you share an split file for this?