This project has a slightly different architecture to a regular deep learning project deployed as an app. We will have two different sets of samples of music. The first set of samples will be used to train an LSTM model that can generate music. The other set of samples will be used as a random input to the LSTM model, which will output the generated music samples. The LSTM-based model that we'll be developing and using later will be deployed on Google Cloud Platform (GCP). You can, however, deploy it on AWS or any other hosting of your choice.
The interaction between the different components that will be used in this project have been summarized in the following diagram:
The mobile application asks the model deployed on the server to generate a new music sample. The model uses a random music sample as input to...
