I have created some pipelines for classification task and I want to check out what information is being present/stored at each stage (e.g. text_stats, ngram_tfidf). How could I do this. pipeline =