Now that we understand how RNNs make for powerful tools of music generation, we'll use the Drums RNN model to do just that. The pre-trained models in Magenta are a good way of starting music generation straightaway. For the Drums RNN model, we'll be using the drum_kit pre-trained bundle, which was trained on thousands of percussion MIDI files.
This section will provide insight into the usage of Magenta on the command line. We'll be primarily using Python code to call Magenta, but using the command line has some advantages:
- It is simple to use and useful for quick use cases.
- It doesn't require writing any code or having any programming knowledge.
- It encapsulates parameters in helpful commands and flags.
In this section, we'll use the Drums RNN model in the command line and learn to configure the generation though...