Abstract
Numerous open-source tools have been introduced to develop and train machine learning models. Compared to other toolkits, TensorFlow attracts the greatest number of users on GitHub. TensorFlow 2 is an improved version of TensorFlow 1 that focuses on simple execution, easy model building, and robust model deployment. This chapter introduces in detail three main updated functions of TensorFlow 2: (1) eager execution, providing a programming environment where operations can be evaluated immediately without constructing graphs; (2) Keras, a high-level application programming interface (API) for quickly and easily designing and training network models; and (3) tf.data, an API that allows users to construct complex input pipelines in a simple way.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
