The idea is to have a trained model on the back-end and generate short looped carol melodies which "make sense" when requested (and later store them in the db).
Basically, all the generation part that is now hard-coded in the front-end should be moved to the back-end and the model should be evaluated and, if necessary, trained again.
The idea is to have a trained model on the back-end and generate short looped carol melodies which "make sense" when requested (and later store them in the db).
Basically, all the generation part that is now hard-coded in the front-end should be moved to the back-end and the model should be evaluated and, if necessary, trained again.