During May 2025, this developer enhanced the pytorch/torchrec repository by implementing an optional R-squared metric for Mean Squared Error (MSE) evaluations, expanding the analytical capabilities available during model assessment. The work involved designing and integrating the necessary computations and state management to support opt-in R-squared calculations within the evaluation pipeline. Using Python and leveraging data analysis and machine learning expertise, the developer ensured that users could access richer evaluation metrics for improved model selection and governance. No major bugs were addressed during this period, with the primary focus on delivering this feature to support more comprehensive model evaluation workflows.