Kangas is an open-source software supplied by Comet, supplier of an MLOps platform for machine studying groups.
Based on Comet, the software is meant to assist groups discover, analyze, and visualize large-scale multimedia knowledge.
Kangas provides customers a Python API for logging giant tables of information in addition to a visible interface for performing sophisticated queries towards the customers dataset.
“A key part of data-centric Machine Studying is having the ability to perceive how your coaching knowledge impacts mannequin outcomes and the place your mannequin predictions are improper,” mentioned Gideon Mendels, CEO and co-founder of Comet. “Kangas accomplishes each of those targets and dramatically improves the expertise for ML practitioners.”
With Kangas, prospects achieve a number of advantages equivalent to scalability with Kangas DataGrid, the category for representing datasets and the power to rapidly group, kind, and filter throughout a number of knowledge factors with a easy UI.
Moreover, this open-source software brings customers interoperability by enabling operating in a pocket book or as a standalone app, each regionally and remotely.
Lastly, this providing brings customers built-in laptop imaginative and prescient help to allow them to visualize and filter bounding containers, labels, and metadata with none extra setup.
For a reside demo of Kangas, click on right here.