SQLite as a Vector Store with SQLiteVec

This notebook covers how to get started with the SQLiteVec vector store.

SQLite-Vec is an SQLite extension designed for vector search, emphasizing local-first operations and easy integration into applications without external servers. It is the successor to SQLite-VSS by the same author. It is written in zero-dependency C and designed to be easy to build and use.

This notebook shows how to use the SQLiteVec vector database.

Setup

You'll need to install langchain-community with pip install -qU langchain-community to use this integration

# You need to install sqlite-vec as a dependency.
%pip install --upgrade --quiet  sqlite-vec

Credentials

SQLiteVec does not require any credentials to use as the vector store is a simple SQLite file.

Initialization

from langchain_community.embeddings.sentence_transformer import (
    SentenceTransformerEmbeddings,
)
from langchain_community.vectorstores import SQLiteVec

embedding_function = SentenceTransformerEmbeddings(model_name="all-MiniLM-L6-v2")
vector_store = SQLiteVec(
    table="state_union", db_file="/tmp/vec.db", embedding=embedding_function
)

API Reference:SentenceTransformerEmbeddings | SQLiteVec

Manage vector store

Add items to vector store

vector_store.add_texts(texts=["Ketanji Brown Jackson is awesome", "foo", "bar"])

Update items in vector store

Not supported yet

Delete items from vector store

Not supported yet

Query vector store

Query directly

data = vector_store.similarity_search("Ketanji Brown Jackson", k=4)

Query by turning into retriever

Not supported yet

Usage for retrieval-augmented generation

Refer to the documentation on sqlite-vec at https://alexgarcia.xyz/sqlite-vec/ for more information on how to use it for retrieval-augmented generation.

API reference

For detailed documentation of all SQLiteVec features and configurations head to the API reference: https://python.langchain.com/api_reference/community/vectorstores/langchain_community.vectorstores.sqlitevec.SQLiteVec.html

Other examples

from langchain_community.document_loaders import TextLoader
from langchain_community.embeddings.sentence_transformer import (
    SentenceTransformerEmbeddings,
)
from langchain_community.vectorstores import SQLiteVec
from langchain_text_splitters import CharacterTextSplitter

# load the document and split it into chunks
loader = TextLoader("../../how_to/state_of_the_union.txt")
documents = loader.load()

# split it into chunks
text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
docs = text_splitter.split_documents(documents)
texts = [doc.page_content for doc in docs]


# create the open-source embedding function
embedding_function = SentenceTransformerEmbeddings(model_name="all-MiniLM-L6-v2")


# load it in sqlite-vss in a table named state_union.
# the db_file parameter is the name of the file you want
# as your sqlite database.
db = SQLiteVec.from_texts(
    texts=texts,
    embedding=embedding_function,
    table="state_union",
    db_file="/tmp/vec.db",
)

# query it
query = "What did the president say about Ketanji Brown Jackson"
data = db.similarity_search(query)

# print results
data[0].page_content

API Reference:TextLoader | SentenceTransformerEmbeddings | SQLiteVec | CharacterTextSplitter

'Tonight. I call on the Senate to: Pass the Freedom to Vote Act. Pass the John Lewis Voting Rights Act. And while you’re at it, pass the Disclose Act so Americans can know who is funding our elections. \n\nTonight, I’d like to honor someone who has dedicated his life to serve this country: Justice Stephen Breyer—an Army veteran, Constitutional scholar, and retiring Justice of the United States Supreme Court. Justice Breyer, thank you for your service. \n\nOne of the most serious constitutional responsibilities a President has is nominating someone to serve on the United States Supreme Court. \n\nAnd I did that 4 days ago, when I nominated Circuit Court of Appeals Judge Ketanji Brown Jackson. One of our nation’s top legal minds, who will continue Justice Breyer’s legacy of excellence.'

Example using existing SQLite connection

from langchain_community.document_loaders import TextLoader
from langchain_community.embeddings.sentence_transformer import (
    SentenceTransformerEmbeddings,
)
from langchain_community.vectorstores import SQLiteVec
from langchain_text_splitters import CharacterTextSplitter

# load the document and split it into chunks
loader = TextLoader("../../how_to/state_of_the_union.txt")
documents = loader.load()

# split it into chunks
text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
docs = text_splitter.split_documents(documents)
texts = [doc.page_content for doc in docs]


# create the open-source embedding function
embedding_function = SentenceTransformerEmbeddings(model_name="all-MiniLM-L6-v2")
connection = SQLiteVec.create_connection(db_file="/tmp/vec.db")

db1 = SQLiteVec(
    table="state_union", embedding=embedding_function, connection=connection
)

db1.add_texts(["Ketanji Brown Jackson is awesome"])
# query it again
query = "What did the president say about Ketanji Brown Jackson"
data = db1.similarity_search(query)

# print results
data[0].page_content

API Reference:TextLoader | SentenceTransformerEmbeddings | SQLiteVec | CharacterTextSplitter

'Ketanji Brown Jackson is awesome'

Related

• Vector store conceptual guide
• Vector store how-to guides