What you'll learn

Ad-hoc data loading and preprocessing often become bottlenecks as models and datasets grow. In this course, Designing Data Pipelines with TensorFlow, you'll gain the ability to build production-ready tf.data pipelines. First, you'll explore the tf.data API and dataset fundamentals, including creating datasets from multiple sources and basic operations. Next, you'll discover preprocessing and transformation pipelines, performance optimization with prefetch and cache, and data augmentation techniques. Finally, you'll learn how to work with large-scale and distributed data using TFRecords, sharding, and tf.distribute. When you're finished with this course, you'll have the skills and knowledge of tf.data pipeline design needed to build efficient, scalable data pipelines for TensorFlow.