Brad Miro, Google
gtc-dc 2019
We’ll define deep learning, which has prompted significant improvements in data science, and is the reason for many advances in the fields of natural language processing, computer vision, and generative modeling. We’ll explain why it might be preferable to machine learning methods, and how to begin building deep learning models using TensorFlow 2.0. We’ll also demonstrate how to use TensorFlow 2.0 to provide native support to scale across multiple GPUs.