This chapter details a machine learning training pipeline to build, train, and validate state-of-the-art machine learning models, including deep neural networks. It describes how to integrate input data pipelines, create tf.keras-based models, perform training in a distributed manner, and run validations to fine-tune model's hyperparameters. It also touches on various concepts on how to export and save TensorFlow models for deployment and inferencing. Model debugging and visualization are the key tools used to debug and improve model accuracy and performance. This chapter also outlines the usage of TensorBoard, changes to it in TF 2.0, and how to use TensorBoard for model debugging and profiling a model's speed and performance.
TensorFlow 1.x version has strong support for low- and mid-level APIs to build machine learning models...
