What are modules?
Modules are the building blocks of your workflow. The three supported modules - Dataset, Augmentations, and Model - represent distinct portions of your computer vision model pipeline. Simply drag and drop the modules (building blocks) to create a workflow.
The workflow follows a top-to-bottom approach, whereby the first module to be run is located at the top of the canvas and final module to be run is located at the bottom of the canvas. You will see an input port and output port located at the top and bottom of the module, which can be thought of as the connecting points between two modules.
Note that not all modules have both input and output ports, this is because some modules must go first or last. For instance, the Dataset Module does not have an input port as it must be the first 'block' of the workflow.
You can also customize each of these modules to define how you would like to train your model (i.e. setting your training parameters). For more details, refer to Module : Dataset, Module : Augmentations, and Module : Model.
Updated about 1 year ago