AI Modules

AI modules are the building blocks of the platform. They are AI models devloped by users to solve particular usecases. And of course, they are open-source! This means that every model you develop in the platform can be deployed anywhere else.

Modules can be explored interactively from the Dashboard, their source code is in Github and the corresponding images are hosted in DockerHub.

ℹ️ Module naming conventions

Modules follow the following name convention:

  • ai4os-hub/ai4-<project-name>: module officially developed by the project

  • ai4os-hub/<project-name>: modules developed by external users

Docker images have usually tags depending on whether they are using Github’s master or test and whether they use cpu or gpu. Tags are usually:

  • latest or cpu: master + cpu

  • gpu: master + gpu

  • cpu-test: test + cpu

  • gpu-test: test + gpu

The following documentation will help you delve deeper into the different aspects of the modules:

 Template

Learn how to fast track your module development starting for our AI module template.

AI Modules Template
 Metadata

Learn how to accurately describe your module for an enhanced discoverability.

AI modules metadata
 Provenance

Learn how our comprehensive provenance chain allows to track how a module was trained.

AI module provenance
 CI/CD pipeline

Learn how to the AI4EOSC CI/CD pipeline enforces best practices in code development.

AI Modules