Apps are programs and tools that run in an AI Workbench project. They are essential components that help you carry out various tasks and operations within your project.
There are three classes of apps in AI Workbench: Web Apps, Processes, and Native Apps apps. Each class has unique characteristics and functions.
- Web Apps
- Processes
- Native Apps
These apps run in the project container and are routed over the reverse proxy. Examples of web apps include JupyterLab, Gradio, and a REST or inference API.
These are programs or scripts that run in the project container without exposing a port or being routed through the proxy. Model downloading scripts are a good example of process class apps.
Native apps are a special class of app that uses an application installed on your host, not in the project container. The only currently supported native app is VS Code.
AI Workbench offers both built-in apps and custom apps. Built-in apps can be added to any project with a single click, while custom apps require you to fill out information to start, stop, and use the app.
Although custom apps are more complicated to configure, they allow you to do anything you want inside a project. They offer greater flexibility and control over your workflow.