You have a text corpus stored in corpus variable. Your task is to display the vector for the 'graphic design' bigram in a BoW model. To do this:

1. Import the CountVectorizer class to create a BoW model.
2. Instantiate the CountVectorizer class as count_vectorizer, configuring it for a frequency-based model that includes both unigrams and bigrams.
3. Use the appropriate method of count_vectorizer to generate a BoW matrix from the 'Document' column in the corpus and store the result in bow_matrix.
4. Convert bow_matrix to a dense array and create a DataFrame from it, setting the unique features (unigrams and bigrams) as its columns. Store the result in the bow_df variable.
5. Display the vector for 'graphic design' bigram as an array.