Skip to main content

Python Code

Python Code & Auto-Containers - a game-changing feature on the Rhino Health FCP. With Auto-Containers, users can effortlessly create working models by providing a .py file and a requirements.txt file, eliminating the need to build or push container images manually. For simple data transformation and preparation tasks using Python code, the Python Code Object offers distributed code execution without the hassle of container management, streamlining the data analysis process for seamless and efficient AI projects.

Generalized Compute

The Generalized Compute (GC) capability offered within the Rhino Health User Guide empowers users to run arbitrary code on one or more remote Datasets through the Rhino Health Federated Computing Platform (FCP). This powerful feature allows data analysts and researchers to interact with input cohorts, execute custom code within containers, and generate new output Datasets, all while ensuring data privacy and security. Whether you choose to containerize your code or run simple Python code, the Generalized Compute functionality simplifies data analysis and AI projects on the Rhino Health FCP, opening doors to a wide range of applications in the healthcare domain.

NVFlare Code

The Rhino Health Federated Computing Platform (FCP) supports federated training of models using the open-source NVIDIA FLARE framework. With NVFlare Code, users can leverage the power of NVIDIA FLARE to perform efficient and collaborative federated training across the platform. This documentation provides comprehensive guidance on how to perform federated training with NVIDIA FLARE and the Rhino Health FCP, ensuring seamless integration and enhanced capabilities for AI projects.

Interactive Containers

In the Rhino Health User Guide, Interactive Containers offer a powerful capability to deploy code with interactive user interfaces (e.g., Jupyter notebooks, image annotation tools) to remote project sites. By preparing a custom Docker image containing preferred software, users can conveniently run it on any participating site and interact with the GUI interface. With direct access to input Datasets loaded into the Docker container, users can seamlessly work with data and create personalized analyses, making it an ideal choice for diverse research and analysis needs. To get started with Interactive Containers, simply create a new Code Object, select the "Interactive Container" Code type, specify your Docker container image, and begin exploring and analyzing your data interactively.