The Music Generation Model is based on Long short-term memory (LSTM), a recurrent neural network (RNN) architecture, that can efficiently learn via gradient descent.
The model is able to generate good quality instrumental music.
WEBSITE LINK: https://music-generation-using-ml.herokuapp.com
$ sudo apt install git
$ git clone https://github.com/ShobhitBansal/Music-generation-Using-Deep-Learning.git
$ cd Music-generation-Using-Deep-Learning
$ sudo apt-get install virtualenv
$ virtualenv env
$ source env/bin/activate
$ sudo apt-get install python3.7
$ sudo apt-get install python3-pip
$ pip3 install -r requirements.txt
$ cd home/model/
$ wget https://www.dropbox.com/s/nn9975383dee3i6/piano_model.hdf5?dl=0 -O piano_model.hdf5
get back to main directory by executing next command 2 times
$ cd ..
$ cd ..
to launch server execute the following
$ python3 app.py
go to Google Chrome and open http://127.0.0.1:5000
Click on the Generate Music button to generate instrumental piano music. The output file will get downloaded as Output.mid. Use any music player to play the music.
