This dataset is a combination of kmfoda/booksum, samsum, mosaicml/dolly_hhrlhf and yahma/alpaca-cleaned.