By default max_df is 1.0, which means “ignore terms that appear in more than 100% of the documents”, while min_df=5 means “ignore terms that appear in more than 5 documents.