The This page can now be found at Sample Text Labeller widget allows to do an inference test to assert the performance of a given Machine Learning Workflow.
It works by providing a sample text that is going to be run through a preselected ML Workflow and the labels that result from that will be shown in the widget. This is useful to test how well a model is performing when labelling Squirro items.
The widget above is configured to use a forward looking classifier, and given that sample text outputs a 'pos' (positive) label with 1.00 confidence score.Widget on the Squirro Docs site.