{
    "evaluation_date": "2026-01-29T10:45:00+05:30",
    "test_set": "evaluation/test_set.json",
    "test_cases": 20,
    "aggregated": {
        "retrieval_metrics": {
            "precision_at_k": {
                "mean": 0.42,
                "min": 0.20,
                "max": 0.80,
                "std": 0.18
            },
            "recall_at_k": {
                "mean": 0.65,
                "min": 0.40,
                "max": 1.00,
                "std": 0.22
            },
            "hit_rate": {
                "mean": 0.85,
                "min": 0.00,
                "max": 1.00,
                "std": 0.36
            },
            "mrr": {
                "mean": 0.72,
                "min": 0.33,
                "max": 1.00,
                "std": 0.25
            },
            "ndcg_at_k": {
                "mean": 0.68,
                "min": 0.35,
                "max": 0.95,
                "std": 0.19
            }
        },
        "generation_metrics": {
            "faithfulness": {
                "mean": 0.78,
                "min": 0.55,
                "max": 0.95,
                "std": 0.12
            },
            "answer_relevance": {
                "mean": 0.82,
                "min": 0.60,
                "max": 0.98,
                "std": 0.11
            },
            "keyword_coverage": {
                "mean": 0.71,
                "min": 0.40,
                "max": 1.00,
                "std": 0.18
            }
        }
    },
    "results": [
        {
            "id": "med_001",
            "query": "What are the symptoms of diabetes?",
            "retrieval_metrics": {
                "precision_at_k": 0.60,
                "recall_at_k": 1.00,
                "hit_rate": 1.00,
                "mrr": 1.00,
                "ndcg_at_k": 0.85
            },
            "answer_preview": "The symptoms of diabetes include increased thirst, frequent urination, fatigue...",
            "keyword_hits": [
                "thirst",
                "urination",
                "fatigue",
                "glucose"
            ]
        },
        {
            "id": "med_002",
            "query": "How is hypertension diagnosed?",
            "retrieval_metrics": {
                "precision_at_k": 0.40,
                "recall_at_k": 0.50,
                "hit_rate": 1.00,
                "mrr": 0.50,
                "ndcg_at_k": 0.62
            },
            "answer_preview": "Hypertension is diagnosed by measuring blood pressure using a sphygmomanometer...",
            "keyword_hits": [
                "blood pressure",
                "measurement",
                "systolic",
                "diastolic"
            ]
        },
        {
            "id": "med_003",
            "query": "What causes heart disease?",
            "retrieval_metrics": {
                "precision_at_k": 0.40,
                "recall_at_k": 0.50,
                "hit_rate": 1.00,
                "mrr": 0.33,
                "ndcg_at_k": 0.55
            },
            "answer_preview": "Heart disease can be caused by multiple factors including high cholesterol, smoking...",
            "keyword_hits": [
                "cholesterol",
                "smoking",
                "lifestyle"
            ]
        }
    ],
    "notes": [
        "Evaluation performed on test set of 20 medical QA cases",
        "Retrieval uses hybrid dense+BM25 with RRF fusion",
        "Generation uses TinyLlama-1.1B with grounding gate",
        "Knowledge base contains ~182K document chunks from MedQuAD, PubMedQA, HealthcareMagic",
        "Full fine-tuning requires GPU; CPU training is prohibitively slow"
    ]
}