About pretrain dataset

[Yang-Liz]

I noticed that you performed preprocessing on a large number of public datasets during the pretraining phase. Will these pretrain datasets be made publicly available later? Alternatively, would you consider publishing the data preprocessing pipeline?