Model checkpoints for models of the WASPAA 2025 paper "Towards Reliable Objective Evaluation Metrics for Generative Singing Voice Separation"