Understanding the TensorFlow deep learning library

Initially released in 2015, TensorFlow is a popular open source machine learning library, primarily backed up by Google, that is mainly designed for deep learning. TensorFlow has been used by companies of all sizes for training and building state-of-the-art deep learning models for a range of use cases, including computer vision, speech recognition, question-answering, text summarization, forecasting, and robotics.

TensorFlow is based on the concept of a computational graph (that is, a dataflow graph), in which the data flow and operations that are performed on the data are constructed as a graph. TensorFlow takes input data in the form of an n-dimensional array/matrix, which is known as a tensor, and performs mathematical operations on this tensor, such as add or matrix multiplication. An example of a tensor could be a scalar value (for example, 1.0), a one-dimensional vector (for example, [1.0, 2.0, 3.0]), a two-dimensional...