In the previous section, we've seen how much we can already do on the command line with the Drums RNN model. In this section, you'll get to create a small application that will use that model to generate music in Python.
Using Magenta in Python is a bit difficult because of the following reasons:
- It requires you to write code and understand Magenta's architecture.
- It requires more boilerplate code and is less straightforward.
But it also has advantages that we think are important:
- You have more freedom in the usage of the models.
- You can create new models and modify existing ones.
- You can go beyond generating single sequences.
The last point is important for us because we'll be building a small music application that generates music autonomously. Calling Magenta's scripts on the command line is convenient, but you cannot...
