An end-to-end benchmark for Gurbani Kirtan alignment and captioning. This project provides a framework to evaluate systems that identify and track specific Shabad lines in live audio streams, snapping predictions to canonical Gurmukhi text rather than using raw ASR.