Fix factual correctness returning only true/false because score variable was overshadowed

Author: rogeriochaves

evaluators/ragas/langevals_ragas/factual_correctness.py

@@ -61,9 +61,7 @@ class RagasFactualCorrectnessEvaluator(
     category = "quality"
     env_vars = env_vars
     default_settings = RagasFactualCorrectnessSettings()
-    docs_url = (
-        "https://docs.ragas.io/en/stable/concepts/metrics/available_metrics/factual_correctness/"
-    )
+    docs_url = "https://docs.ragas.io/en/stable/concepts/metrics/available_metrics/factual_correctness/"
     is_guardrail = False
 
     def evaluate(self, entry: RagasFactualCorrectnessEntry) -> SingleEvaluationResult:
@@ -121,16 +119,16 @@ async def verify_claims(premise: str, hypothesis_list: list[str], callbacks):
             details += (
                 f"# Precision\nPremise: {breakdown['premise']}\nHypothesis list:\n"
             )
-            for i, score in enumerate(breakdown["scores"]):
-                details += f"- \"{breakdown['hypothesis_list'][i]}\": {score}\n"
+            for i, score_ in enumerate(breakdown["scores"]):
+                details += f"- \"{breakdown['hypothesis_list'][i]}\": {score_}\n"
 
         if len(breakdowns) > 1:
             breakdown = breakdowns[1]
             details += (
                 f"\n# Recall\nPremise: {breakdown['premise']}\nHypothesis list:\n"
             )
-            for i, score in enumerate(breakdown["scores"]):
-                details += f"- \"{breakdown['hypothesis_list'][i]}\": {score}\n"
+            for i, score_ in enumerate(breakdown["scores"]):
+                details += f"- \"{breakdown['hypothesis_list'][i]}\": {score_}\n"
 
         return RagasResult(
             score=score,

evaluators/ragas/tests/test_ragas.py

@@ -72,28 +72,10 @@ def test_faithfulness():
     result = evaluator.evaluate(
         RagasFaithfulnessEntry(
             output="The capital of France is Paris.",
-            contexts=["France is a country in Europe.", "Paris is a city in France."],
-        )
-    )
-
-    assert result.status == "processed"
-    assert result.score and result.score > 0.9
-    assert result.cost and result.cost.amount > 0.0
-    assert result.details
-
-
-@pytest.mark.flaky(max_runs=3)
-def test_faithfulness_hhem():
-    evaluator = RagasFaithfulnessEvaluator(
-        settings=RagasFaithfulnessSettings(use_hhem=True, model="openai/gpt-3.5-turbo")
-    )
-
-    result = evaluator.evaluate(
-        RagasFaithfulnessEntry(
-            input="When was the first super bowl?",
-            output="The first superbowl was held on Jan 15, 1967",
             contexts=[
-                "The First AFL–NFL World Championship Game was an American football game played on January 15, 1967, at the Los Angeles Memorial Coliseum in Los Angeles."
+                "France is a country in Europe.",
+                "Paris is a city in France.",
+                "Paris is the capital of France.",
             ],
         )
     )
@@ -373,8 +355,6 @@ def test_rubrics_based_scoring_without_reference():
         )
     )
 
-    print("\n\nresult", result, "\n\n")
-
     assert result.status == "processed"
     assert result.score and result.score == 5
     assert result.cost and result.cost.amount > 0.0