Test Set Diagnostics
Model generalization performance, error distributions, and calibration curves analyzed on unseen test samples.
Training Set Diagnostics
Evaluation diagnostics on the training subset, useful for monitoring convergence and detecting overfitting.