Create VectorStore interface (#92)

examples/embeddings.ipynb

@@ -8,9 +8,9 @@
    "outputs": [],
    "source": [
     "from langchain.embeddings.openai import OpenAIEmbeddings\n",
-    "from langchain.elastic_vector_search import ElasticVectorSearch\n",
+    "from langchain.text_splitter import CharacterTextSplitter\n",
-    "from langchain.faiss import FAISS\n",
+    "from langchain.vectorstores.elastic_vector_search import ElasticVectorSearch\n",
-    "from langchain.text_splitter import CharacterTextSplitter"
+    "from langchain.vectorstores.faiss import FAISS"
    ]
   },
   {
@@ -69,7 +69,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 5,
    "id": "4906b8a3",
    "metadata": {},
    "outputs": [],
@@ -82,7 +82,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 6,
    "id": "95f9eee9",
    "metadata": {},
    "outputs": [

langchain/__init__.py

@@ -16,11 +16,10 @@ from langchain.chains import (
     SQLDatabaseChain,
 )
 from langchain.docstore import Wikipedia
-from langchain.elastic_vector_search import ElasticVectorSearch
-from langchain.faiss import FAISS
 from langchain.llms import Cohere, HuggingFaceHub, OpenAI
 from langchain.prompts import BasePrompt, DynamicPrompt, Prompt
 from langchain.sql_database import SQLDatabase
+from langchain.vectorstores import FAISS, ElasticVectorSearch
 
 __all__ = [
     "LLMChain",

langchain/vectorstores/__init__.py

@@ -0,0 +1,6 @@
+"""Wrappers on top of vector stores."""
+from langchain.vectorstores.base import VectorStore
+from langchain.vectorstores.elastic_vector_search import ElasticVectorSearch
+from langchain.vectorstores.faiss import FAISS
+
+__all__ = ["ElasticVectorSearch", "FAISS", "VectorStore"]

langchain/vectorstores/base.py

@@ -0,0 +1,13 @@
+"""Interface for vector stores."""
+from abc import ABC, abstractmethod
+from typing import List
+
+from langchain.docstore.document import Document
+
+
+class VectorStore(ABC):
+    """Interface for vector stores."""
+
+    @abstractmethod
+    def similarity_search(self, query: str, k: int = 4) -> List[Document]:
+        """Return docs most similar to query."""

langchain/vectorstores/elastic_vector_search.py

@@ -4,6 +4,7 @@ from typing import Callable, Dict, List
 
 from langchain.docstore.document import Document
 from langchain.embeddings.base import Embeddings
+from langchain.vectorstores.base import VectorStore
 
 
 def _default_text_mapping(dim: int) -> Dict:
@@ -27,7 +28,7 @@ def _default_script_query(query_vector: List[int]) -> Dict:
     }
 
 
-class ElasticVectorSearch:
+class ElasticVectorSearch(VectorStore):
     """Wrapper around Elasticsearch as a vector database.
 
     Example:

langchain/vectorstores/faiss.py

@@ -7,9 +7,10 @@ from langchain.docstore.base import Docstore
 from langchain.docstore.document import Document
 from langchain.docstore.in_memory import InMemoryDocstore
 from langchain.embeddings.base import Embeddings
+from langchain.vectorstores.base import VectorStore
 
 
-class FAISS:
+class FAISS(VectorStore):
     """Wrapper around FAISS vector database.
 
     To use, you should have the ``faiss`` python package installed.