Generation of music based on 7 fundamental modes, other modes are (and can be) added as well. Steps:

1. Collect the songs/datsets
2. Identify the key and mode for each songs
3. Transpose all the keys to a reference key of C
4. Make separate datasets for each modes in the key of C
5. Train the models for each musical modes
6. Prompt the composer to provide inputs
7. Generate a sequence of notes/melody
8. Reverse transpose the notes based on the prompt
9. Play the melody to the user
10. Take feedback