Attendees will learn how to utilize deep learning techniques to process remote sensing (RS) data in a 4-hour workshop. The first three hours will be dedicated to learning PyTorch (and other Python tools), an open-source deep learning development platform. Participants will be introduced to the fundamentals of deep learning, learn how to create deep learning models from scratch, and explore customized applications such as image classification using PyTorch Image Models (TIMM) and image segmentation with Segmentation Models PyTorch (SMP). The final hour will focus on 3D RS data generation and processing. Using PyTorch3D, we will introduce the implementation of the structure-from-motion (SfM) algorithm based on 2D images to generate georeferenced point clouds. With brief exposure to digital surface and terrain model generation using point cloud or LiDAR data, advanced semantic 3D classification and segmentation models will be introduced. Upon completion of the workshop, attendees will possess the necessary skills to develop customized applications using Python tools for advanced RS data processing.