rag_evaluation_report.json

{
  "dataset_size": 5,
  "frameworks": {
    "custom_retrieval": {
      "context_precision": 1.0,
      "context_recall": 1.0,
      "noise_rejection": 1.0,
      "mrr": 1.0
    },
    "custom_generation": {
      "rouge_l": 0.7554,
      "faithfulness": 0.7333,
      "answer_relevancy": 0.8779,
      "verbatim_overlap": 0.4422,
      "hallucination_free": 1.0
    },
    "rouge": {
      "status": "completed",
      "average_scores": {
        "rouge1": 0.785,
        "rouge2": 0.7231,
        "rougeL": 0.7554
      },
      "per_case": {
        "E4-main-walls-good": {
          "rouge1": 0.8235,
          "rouge2": 0.7711,
          "rougeL": 0.8235
        },
        "E5-roof-defects": {
          "rouge1": 0.878,
          "rouge2": 0.825,
          "rougeL": 0.878
        },
        "E7-windows": {
          "rouge1": 0.8235,
          "rouge2": 0.7952,
          "rougeL": 0.8235
        },
        "D-property-overview": {
          "rouge1": 0.6667,
          "rouge2": 0.5,
          "rougeL": 0.5185
        },
        "F1-garage": {
          "rouge1": 0.7333,
          "rouge2": 0.7241,
          "rougeL": 0.7333
        }
      }
    },
    "ragas": {
      "status": "completed",
      "scores": {
        "faithfulness": 0.5467,
        "answer_relevancy": 0.781,
        "llm_context_precision_with_reference": 0.8667,
        "context_recall": 0.5333
      }
    },
    "deepeval": {
      "status": "completed",
      "scores": {
        "faithfulness": 1.0,
        "answer_relevancy": 1.0,
        "contextual_precision": 1.0,
        "contextual_recall": 0.6167,
        "hallucination": 0.0
      },
      "per_case": {
        "faithfulness": [
          1.0,
          1.0,
          1.0,
          1.0,
          1.0
        ],
        "answer_relevancy": [
          1.0,
          1.0,
          1.0,
          1.0,
          1.0
        ],
        "contextual_precision": [
          1.0,
          1.0,
          1.0,
          1.0,
          1.0
        ],
        "contextual_recall": [
          0.75,
          0.6666666666666666,
          0.5,
          0.5,
          0.6666666666666666
        ],
        "hallucination": [
          0.0,
          0.0,
          0.0,
          0.0,
          0.0
        ]
      }
    }
  }
}