Comment on: Empromptu AI

Most fine tuning tools treat evaluation as an afterthought does Empromptu have a built in framework for measuring whether a fine tuned model is actually outperforming the base model in production not just on a held out test set?