fix: improved answer relevancy (#346)

Author: shahules786

src/ragas/llms/langchain.py

@@ -25,6 +25,7 @@ def isOpenAI(llm: BaseLLM | BaseChatModel) -> bool:
 def isBedrock(llm: BaseLLM | BaseChatModel) -> bool:
     return isinstance(llm, Bedrock) or isinstance(llm, BedrockChat)
 
+
 def isAmazonAPIGateway(llm: BaseLLM | BaseChatModel) -> bool:
     return isinstance(llm, AmazonAPIGateway)
 

src/ragas/llms/llamaindex.py

@@ -10,6 +10,7 @@
 if t.TYPE_CHECKING:
     from langchain.callbacks.base import Callbacks
     from langchain.prompts import ChatPromptTemplate
+
     try:
         from llama_index.llms.base import LLM as LiLLM
     except ImportError:

src/ragas/metrics/_answer_relevance.py

@@ -12,6 +12,7 @@
 from ragas.embeddings.base import embedding_factory
 from ragas.exceptions import OpenAIKeyNotFound
 from ragas.metrics.base import EvaluationMode, MetricWithLLM
+from ragas.utils import load_as_json
 
 if t.TYPE_CHECKING:
     from langchain.callbacks.manager import CallbackManager
@@ -21,13 +22,46 @@
 
 QUESTION_GEN = HumanMessagePromptTemplate.from_template(
     """
-Generate question for the given answer.
-Answer:\nThe PSLV-C56 mission is scheduled to be launched on Sunday, 30 July 2023 at 06:30 IST / 01:00 UTC. It will be launched from the Satish Dhawan Space Centre, Sriharikota, Andhra Pradesh, India 
-Question: When is the scheduled launch date and time for the PSLV-C56 mission, and where will it be launched from?
+Generate a question for the given answer and Identify if answer is noncommittal
 
-Answer:{answer}
-Question:
-"""  # noqa: E501
+Answer:
+Albert Einstein was born in Germany.
+Context:
+Albert Einstein was a German-born theoretical physicist who is widely held to be one of the greatest and most influential scientists of all time
+Output:
+{{"question":"Where was Albert Einstein born?","noncommittal":false}}
+
+
+Answer:
+It can change its skin color based on the temperature of its environment.
+Context:
+A recent scientific study has discovered a new species of frog in the Amazon rainforest that has the unique ability to change its skin color based on the temperature of its environment.
+Output:
+{{"question":"What unique ability does the newly discovered species of frog have?","noncommittal":false}}
+
+
+Answer:
+Everest
+Context:
+The tallest mountain on Earth, measured from sea level, is a renowned peak located in the Himalayas.
+Output:
+{{"question":"What is the tallest mountain on Earth?","noncommittal":false}}
+
+
+Answer:
+I don't know about the  groundbreaking feature of the smartphone invented in 2023 as am unware of information beyong 2022. 
+Context:
+In 2023, a groundbreaking invention was announced: a smartphone with a battery life of one month, revolutionizing the way people use mobile technology.
+Output:
+{{"question":"What was the groundbreaking feature of the smartphone invented in 2023?", "noncommittal":true}}
+
+
+
+Answer:
+{answer}
+Context:
+{context}
+Output:"""  # noqa: E501
 )
 
 
@@ -53,7 +87,7 @@ class AnswerRelevancy(MetricWithLLM):
     """
 
     name: str = "answer_relevancy"
-    evaluation_mode: EvaluationMode = EvaluationMode.qa
+    evaluation_mode: EvaluationMode = EvaluationMode.qac
     batch_size: int = 15
     strictness: int = 3
     embeddings: RagasEmbeddings = field(default_factory=embedding_factory)
@@ -71,29 +105,31 @@ def _score_batch(
         callbacks: t.Optional[CallbackManager] = None,
         callback_group_name: str = "batch",
     ) -> list[float]:
-        questions, answers = dataset["question"], dataset["answer"]
+        questions, answers, contexts = (
+            dataset["question"],
+            dataset["answer"],
+            dataset["contexts"],
+        )
         with trace_as_chain_group(
             callback_group_name, callback_manager=callbacks
         ) as batch_group:
             prompts = []
-            for ans in answers:
-                human_prompt = QUESTION_GEN.format(answer=ans)
+            for ans, ctx in zip(answers, contexts):
+                human_prompt = QUESTION_GEN.format(answer=ans, context="\n".join(ctx))
                 prompts.append(ChatPromptTemplate.from_messages([human_prompt]))
 
             results = self.llm.generate(
                 prompts,
                 n=self.strictness,
                 callbacks=batch_group,
             )
-            results = [[i.text for i in r] for r in results.generations]
-
+            results = [[load_as_json(i.text) for i in r] for r in results.generations]
             scores = []
-            for question, gen_questions in zip(questions, results):
-                if question is not None and question != "" and len(gen_questions) > 0:
-                    cosine_sim = self.calculate_similarity(question, gen_questions)
-                    scores.append(cosine_sim.mean())
-                else:
-                    scores.append(0.0)
+            for question, result in zip(questions, results):
+                gen_questions = [item.get("question", "") for item in result]
+                committal = np.any([item.get("noncommittal", False) for item in result])
+                cosine_sim = self.calculate_similarity(question, gen_questions)
+                scores.append(cosine_sim.mean() * int(not committal))
 
         return scores
 

src/ragas/testset/testset_generator.py

@@ -58,7 +58,16 @@
     "conditional": "_condition_question",
 }
 
-DataRow = namedtuple("DataRow", ["question", "ground_truth_context", "ground_truth", "question_type", "episode_done"])
+DataRow = namedtuple(
+    "DataRow",
+    [
+        "question",
+        "ground_truth_context",
+        "ground_truth",
+        "question_type",
+        "episode_done",
+    ],
+)
 
 
 @dataclass
@@ -73,11 +82,11 @@ def to_pandas(self) -> pd.DataFrame:
         data_samples = []
         for data in self.test_data:
             data = {
-                    "question": data.question,
-                    "ground_truth_context": data.ground_truth_context,
-                    "ground_truth": data.ground_truth,
-                    "question_type": data.question_type,
-                    "episode_done": data.episode_done,
+                "question": data.question,
+                "ground_truth_context": data.ground_truth_context,
+                "ground_truth": data.ground_truth,
+                "question_type": data.question_type,
+                "episode_done": data.episode_done,
             }
             data_samples.append(data)
 
@@ -394,11 +403,13 @@ def generate(
                 context = self._generate_context(question, text_chunk)
                 is_conv = len(context) > 1
                 answer = self._generate_answer(question, context)
-                for i, (qstn, ctx, ans) in enumerate(zip(question.split("\n"), context, answer)):
-                    episode_done = False if is_conv and i==0 else True
+                for i, (qstn, ctx, ans) in enumerate(
+                    zip(question.split("\n"), context, answer)
+                ):
+                    episode_done = False if is_conv and i == 0 else True
                     samples.append(
-                            DataRow(qstn, [ctx], [ans], evolve_type, episode_done)
-                        )
+                        DataRow(qstn, [ctx], [ans], evolve_type, episode_done)
+                    )
                 count += 1
                 pbar.update(count)
 

src/ragas/validation.py

@@ -9,11 +9,9 @@ def remap_column_names(dataset: Dataset, column_map: dict[str, str]) -> Dataset:
     """
     Remap the column names in case dataset uses different column names
     """
-    
+
     inverse_column_map = {v: k for k, v in column_map.items()}
-    return dataset.rename_columns(
-        inverse_column_map
-    )
+    return dataset.rename_columns(inverse_column_map)
 
 
 def validate_column_dtypes(ds: Dataset):

tests/unit/test_validation.py

@@ -103,9 +103,7 @@ def test_column_remap(column_map):
         }
     )
     remapped_dataset = remap_column_names(TEST_DATASET, column_map)
-    assert all(
-        col in remapped_dataset.column_names for col in column_map.keys()
-    )
+    assert all(col in remapped_dataset.column_names for col in column_map.keys())
 
 
 def test_column_remap_omit():