Dear Apertus Team, (When) do you plan to release the full dataset for training (similar to how AI2 released Dolma for OLmo?) Best Regards