{ "total_labeled_cases": 97, "total_unlabeled_cases": 179, "1a) QA model answer rate in top 3": 0.92, "1b) QA model okay rate in top 3": 0.08, "1c) QA model bad rate in top 3": 0, "total labeled in top 3": 25, "1a) QA model answer rate in top 5": 0.872, "1b) QA model okay rate in top 5": 0.077, "1c) QA model bad rate in top 5": 0.051, "total labeled in top 5": 39, "2) QA model ndcg metric": 0.919, "3a) QA model bad answer placement error - median": 5, "3b) QA model bad answer placement error - 75th percentile": 9, "4a) QA model answer placement error - median": 2.25, "4b) QA model answer placement error - 75th percentile": 3.286 }