Wow, that's an impressive set of practical examples of using Deep Learning Projects in Python to build solutions in the creative space! Let's revisit the goals we set up for ourselves:
In this project, we're going to take the next step in our computational linguistics journey in Deep Learning Projects in Python and GENERATE new content for our client. We need to help them by providing a deep learning solution that generates new content that can be used in movie script, song lyrics, and music.
Deep Learning generated content for creative purposes is obviously very tricky. Our realistic goal in this chapter was to demonstrate and train you on the skills and architecture needed to get started on these types of projects. Producing acceptable results takes your interaction with the data, the model, the outputs AND testing with appropriate audiences. The key takeaway to remember is that the outputs of your models can be quite personalized to the task and to expand your thinking of what types of business use cases you should feel comfortable working on in your career.
In this chapter, we implemented a generative model which generated content using the LSTM's. We implemented models for both text and audio that generated content for artists and various businesses in the creative space (hypothetically): the Music and Movie industries.
What we learned in this chapter was:
- Text generation with LSTM
- The Additional power of a Bi-directional LSTM for text generation
- Deep (Multi-layer) LSTM to generate lyrics for a song
- Deep (Mulit-layer) LSTM to generate the music for a song
Exciting work in Deep Learning and it keeps on coming in the next chapter...let's see what's in store!