Implemented Reciprocal Rank Fusion (#87)

* Implemented Ensemble Retriever

* Used self.alpha for the weights

* Cleaned up the return statement

sage/vector_store.py

@@ -8,7 +8,8 @@
 
 import marqo
 import nltk
-from langchain_community.retrievers import PineconeHybridSearchRetriever
+from langchain_community.retrievers import BM25Retriever
+from langchain.retrievers import EnsembleRetriever
 from langchain_community.vectorstores import Marqo
 from langchain_community.vectorstores import Pinecone as LangChainPinecone
 from langchain_core.documents import Document
@@ -133,19 +134,23 @@ def upsert_batch(self, vectors: List[Vector], namespace: str):
         self.index.upsert(vectors=pinecone_vectors, namespace=namespace)
 
     def as_retriever(self, top_k: int, embeddings: Embeddings, namespace: str):
-        if self.bm25_encoder:
-            return PineconeHybridSearchRetriever(
-                embeddings=embeddings,
-                sparse_encoder=self.bm25_encoder,
-                index=self.index,
-                namespace=namespace,
-                top_k=top_k,
-                alpha=self.alpha,
-            )
-
-        return LangChainPinecone.from_existing_index(
+        bm25_retriever = BM25Retriever(
+            embeddings=embeddings,
+            sparse_encoder=self.bm25_encoder,
+            index=self.index,
+            namespace=namespace,
+            top_k=top_k,
+        ) if self.bm25_encoder else None
+
+        dense_retriever = LangChainPinecone.from_existing_index(
             index_name=self.index_name, embedding=embeddings, namespace=namespace
         ).as_retriever(search_kwargs={"k": top_k})
+
+        if bm25_retriever:
+            return EnsembleRetriever(retrievers=[dense_retriever, bm25_retriever], weights=[self.alpha, 1-self.alpha])
+        else:
+            return dense_retriever
+
 
 
 class MarqoVectorStore(VectorStore):