Replicate SDD dataset

[Hvidaugust]

Hello!
I am very impressed by your work with this paper, and wanted to use this for comparison for another project, and wanted to use your preprocessed dataset. I tried recreating the dataset using then included SDD script, but find that the generated dataset is much larger than the one found in the zip-file, where the given train-dataset seems to include ~8.5K samples, whereas the generated one includes ~135K.
Am I doing something wrong? Was any kind of filtering done?

Thank you in advance!
Regards E

[22dido22]

Hello @Hvidaugust, is it possible to have your code ? thank you.