Description
Machine-learning metrics can be computed consistently across PyTorch training and evaluation loops. Researchers and ML engineers can track classification, regression, ranking, and distributed metrics at scale. It processes model outputs and labels that may reflect sensitive datasets.