An integration package connecting Cognee and LangChain
Project description
langchain-cognee
This package contains the LangChain integration with cognee.
This package enables you to:
- Ingest documents into cognee
- Build or update a knowledge graph
- Retrieve and query your data using LangChain's standard interfaces
For more information, check out cognee documentation.
Installation
pip install -U langchain-cognee
Configuration
Set your environment variables required by cognee:
export LLM_API_KEY="your-openai-api-key"
Cognee's default settings:
- LLM Provider: OpenAI
- Databases: SQLite, LanceDB, networkx
In case you want to customize your settings, please refer here and configure your env variables accordingly.
Supported databases
- Relational databases: SQLite, PostgreSQL
- Vector databases: LanceDB, PGVector, QDrant, Weviate
- Graph databases: Neo4j, NetworkX
Basic Usage
Below is a minimal example of how to use this integration:
from langchain_cognee.retrievers import CogneeRetriever
from langchain_core.documents import Document
# 1) Instantiate the retriever
retriever = CogneeRetriever(
llm_api_key="YOUR_KEY",
dataset_name="test_dataset",
k=3
)
# 2) (Optional) Reset dataset if you want a clean slate
retriever.reset_dataset()
# 3) Add documents
docs = [
Document(page_content="Elon Musk is the CEO of SpaceX."),
Document(page_content="SpaceX focuses on rockets."),
]
retriever.add_documents(docs)
# 4) Build knowledge graph
retriever.process_data()
# 5) Retrieve documents
results = retriever.invoke("Tell me about Elon Musk")
for doc in results:
print(doc.page_content)
You can also incorporate CogneeRetriever in any LangChain chain.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file langchain_cognee-0.1.1.tar.gz
.
File metadata
- Download URL: langchain_cognee-0.1.1.tar.gz
- Upload date:
- Size: 6.3 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: poetry/2.0.1 CPython/3.11.5 Darwin/23.6.0
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 |
1ab75f6110591779dc11cb63379767524179e0b847eba8507023df951051c100
|
|
MD5 |
c2481f7bf0826d0630580aafca4d4761
|
|
BLAKE2b-256 |
034d985b98a08991ae00bf6ac44ede419cb9b78d6f463f80b03aa44823542cb7
|
File details
Details for the file langchain_cognee-0.1.1-py3-none-any.whl
.
File metadata
- Download URL: langchain_cognee-0.1.1-py3-none-any.whl
- Upload date:
- Size: 9.2 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: poetry/2.0.1 CPython/3.11.5 Darwin/23.6.0
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 |
000d1b21bc3461bf7d6f46e27710563b773ce784641a691524b3b33749e1fbcd
|
|
MD5 |
3ca3733ab1f10af66c0f5ab96bb82bea
|
|
BLAKE2b-256 |
12e06ec0826eeef2d7dd75ba3621464d997a3136e3642a5cc2fe5dcc8c64aeb5
|