Chapter 3. Exploring the Library and the Ecosystem

TensorFlow itself, while impressive, is “just” an open source library for numerical computation using data flow graphs. As described in Chapter 2, there are plenty of open source competitors that you could use to build, train, and run inference with complex neural networks; more will arise. It is the ecosystem surrounding a library, built not only by the original author(s) but also by the community, that forms a long-term solution in an ever-evolving space. It is TensorFlow’s rich and growing ecosystem that compels many to use it. To go through the detailed use of each TensorFlow component would be beyond the scope of this report, but we will strive to introduce the relevant pieces and provide some perspective on the larger overall puzzle.

Python serves as an excellent example of the power of the ecosystem. One of the language’s selling points was its “batteries-included” philosophy; it came with a standard library that made many tasks (such as making HTTP requests) simple. Even with this approach, Python owes some of its success to its ecosystem. The Numpy and Scipy libraries created a strong foundation for numerical and scientific computing, extending the language’s core capabilities and the community that uses it. Libraries such as scikit-learn, which serves almost as a reference implementation for algorithms within the field of machine learning, and Pandas, the de facto standard for Python-based data analysis, have built upon ...

Get Considering TensorFlow for the Enterprise now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.