Live Demo
This is a real, unmodified interactive_pipeline.html produced by the Machine Learning Pipeline example — the iris dataset run through two scalers, three embedding methods, and two clustering algorithms, tracked as 21 nodes across four generations.
Explore it exactly as you would on your own machine: drag to pan, scroll to zoom, and click any node to inspect its metadata — hyperparameters, scores, the cluster plots rendered inline as figures, and links to every artifact the node produced.
Generate your own
This file is the output of a single call — store.generate_web_graph() — and is fully self-contained: no server, no dependencies, just open it in a browser or share it as-is.