# Introduction Automatic differential is very important for machine learning algorithms, especially the backpropagation process when training neural networks. So, TensorFlow 2 provides the gradient tracking process in eager execution mode.
Click the virtual machine below to start practicing