ScalingOpt is a professional platform dedicated to large-scale deep learning optimization, promoting verifiable and scalable algorithms. It aggregates cutting-edge optimizers, provides benchmark analysis, visualization demos, curated literature, and educational tutorials. The community-driven project aims to foster collaboration and advance research in optimization methodology. We welcome contributions and feedback. Your engagement and GitHub stars are greatly appreciated.