Skip to content

Part I: Understanding TensorFlow Fundamentals

“You can’t build AI until you understand the engine that powers it.”

Part I lays the foundation for everything you’ll do with TensorFlow. Whether you’re building a neural network, deploying to mobile, or diving into transformers, it all starts here—with understanding what TensorFlow actually is, how it works under the hood, and how to set it up properly.

By the end of this section, you’ll know how TensorFlow thinks, and how to speak its language fluently—starting with tensors, variables, and graphs.

✅ Chapter 1: What is TensorFlow? An introduction to the TensorFlow ecosystem, its purpose, and its core components. You’ll learn what makes it different from other frameworks and why it’s more than just a library—it's a complete ML platform.

✅ Chapter 2: Architecture of TensorFlow A deeper look into how TensorFlow works internally. You'll explore computation graphs, eager execution, the gradient engine, the deployment stack (TFLite, TFX), and how everything fits together.

✅ Chapter 3: TensorFlow vs Keras Not all Keras is TensorFlow—and not all TensorFlow needs Keras. This chapter clarifies the relationship between them, compares standalone Keras with tf.keras, and shows which one to use when (spoiler: it's tf.keras).

✅ Chapter 4: Installing TensorFlow (Windows, Linux, MacOS) You’ll learn how to properly install TensorFlow with or without GPU support, how to create virtual environments, and how to verify CUDA and device configuration. Practical, essential, and clean.

✅ Chapter 5: First Tensor Example – Hello, tf.Tensor! This is your “hello world” moment. You’ll learn how to create tensors, inspect them, perform basic math, and understand shapes, ranks, and types. It's the first hands-on look at how TensorFlow represents and manipulates data.

After Part I, You Will Be Able To:

  • Understand the role and scope of TensorFlow in machine learning

  • Differentiate between TensorFlow and Keras (and when to use each)

  • Set up your local dev environment (with GPU acceleration)

  • Create and manipulate basic tensors with confidence

  • Prepare for deeper topics like training models and computing gradients

Part I is where the groundwork is laid. The rest of the book builds upon it—one tensor at a time.