Skip to content

lindorm.mdx

Latest commit

 

History

History
147 lines (106 loc) · 4.02 KB

lindorm.mdx

File metadata and controls

147 lines (106 loc) · 4.02 KB
title LindormVectorStore integration
description Integrate with the LindormVectorStore using LangChain Python.

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

Setup

To access Lindorm vector stores you'll need to create a Lindorm account, get the ak/sk, and install the langchain-lindorm-integration integration package.

pip install -qU "langchain-lindorm-integration"

Credentials

Head to here to sign up to Lindorm and generate the ak/sk.

import os


class Config:
    SEARCH_ENDPOINT = os.environ.get("SEARCH_ENDPOINT", "SEARCH_ENDPOINT")
    SEARCH_USERNAME = os.environ.get("SEARCH_USERNAME", "root")
    SEARCH_PWD = os.environ.get("SEARCH_PASSWORD", "<PASSWORD>")
    AI_LLM_ENDPOINT = os.environ.get("AI_ENDPOINT", "<AI_ENDPOINT>")
    AI_USERNAME = os.environ.get("AI_USERNAME", "root")
    AI_PWD = os.environ.get("AI_PASSWORD", "<PASSWORD>")
    AI_DEFAULT_EMBEDDING_MODEL = "bge_m3_model"  # set to your model

Initialization

here we use the embedding model deployed on Lindorm AI Service.

from langchain_lindorm_integration.embeddings import LindormAIEmbeddings
from langchain_lindorm_integration.vectorstores import LindormVectorStore

embeddings = LindormAIEmbeddings(
    endpoint=Config.AI_LLM_ENDPOINT,
    username=Config.AI_USERNAME,
    password=Config.AI_PWD,
    model_name=Config.AI_DEFAULT_EMBEDDING_MODEL,
)

index = "test_index"
vector = embeddings.embed_query("hello word")
dimension = len(vector)
vector_store = LindormVectorStore(
    lindorm_search_url=Config.SEARCH_ENDPOINT,
    embedding=embeddings,
    http_auth=(Config.SEARCH_USERNAME, Config.SEARCH_PWD),
    dimension=dimension,
    embeddings=embeddings,
    index_name=index,
)

Manage vector store

Add items to vector store

from langchain_core.documents import Document

document_1 = Document(page_content="foo", metadata={"source": "https://example.com"})

document_2 = Document(page_content="bar", metadata={"source": "https://example.com"})

document_3 = Document(page_content="baz", metadata={"source": "https://example.com"})

documents = [document_1, document_2, document_3]

vector_store.add_documents(documents=documents, ids=["1", "2", "3"])
['1', '2', '3']

Delete items from vector store

vector_store.delete(ids=["3"])
{'took': 400,
 'timed_out': False,
 'total': 1,
 'deleted': 1,
 'batches': 1,
 'version_conflicts': 0,
 'noops': 0,
 'retries': {'bulk': 0, 'search': 0},
 'throttled_millis': 0,
 'requests_per_second': -1.0,
 'throttled_until_millis': 0,
 'failures': []}

Query vector store

Once your vector store has been created and the relevant documents have been added you will most likely wish to query it during the running of your chain or agent.

Query directly

Performing a simple similarity search can be done as follows:

results = vector_store.similarity_search(query="thud", k=1)
for doc in results:
    print(f"* {doc.page_content} [{doc.metadata}]")
* foo [{'source': 'https://example.com'}]

If you want to execute a similarity search and receive the corresponding scores you can run:

results = vector_store.similarity_search_with_score(query="thud", k=1)
for doc, score in results:
    print(f"* [SIM={score:3f}] {doc.page_content} [{doc.metadata}]")
* [SIM=0.671268] foo [{'source': 'https://example.com'}]

Usage for retrieval-augmented generation

For guides on how to use this vector store for retrieval-augmented generation (RAG), see the following sections:

API reference

For detailed documentation of all LindormVectorStore features and configurations head to the API reference.