Update README.md
Browse files
README.md
CHANGED
@@ -7,4 +7,41 @@ sdk: static
|
|
7 |
pinned: false
|
8 |
---
|
9 |
|
10 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
pinned: false
|
8 |
---
|
9 |
|
10 |
+
# Graph RAG
|
11 |
+
|
12 |
+
Retrievers providing both **unstructured** (similarity-search on vectors) and
|
13 |
+
**structured** (traversal of metadata properties).
|
14 |
+
|
15 |
+
## About The Project
|
16 |
+
|
17 |
+
Graph RAG provides retrievers combining vector-search (for unstructured similarity) and traversal (for structured relationships in metadata).
|
18 |
+
These retrievers are implemented using the metadata search functionality of existing vector stores, **allowing you to traverse your existing vector store**!
|
19 |
+
|
20 |
+
The core library (`graph-retriever`) supports can be used in generic Python applications, while `langchain-graph-retriever` provides [langchain](https://python.langchain.com/docs/introduction/)-specific functionality.
|
21 |
+
|
22 |
+
<!-- GETTING STARTED -->
|
23 |
+
## Getting Started with LangChain
|
24 |
+
|
25 |
+
1. Install `langchain-graph-retriever` (or add to your Python dependencies).
|
26 |
+
|
27 |
+
```sh
|
28 |
+
pip install langchain-graph-retriever
|
29 |
+
```
|
30 |
+
|
31 |
+
1. Wrap your existing vector store to enable graph retrieval:
|
32 |
+
|
33 |
+
```python
|
34 |
+
from langchain_graph_retriever import GraphRetriever
|
35 |
+
|
36 |
+
retriever = GraphRetriever(
|
37 |
+
# Adapt AstraDBVectorStore for use with Graph Retrievers.
|
38 |
+
# Exposes functionality of the underlying store that is otherwise not available.
|
39 |
+
store = store,
|
40 |
+
# Define the relationships to navigate:
|
41 |
+
# 1. From nodes with a list of `mentions` to the nodes with the corresponding `ids`.
|
42 |
+
# 2. From nodes with a list of related `entities` to other nodes with the same entities.
|
43 |
+
edges = [("mentions", "id"), "entities"],
|
44 |
+
)
|
45 |
+
|
46 |
+
retriever.invoke("where is Santa Clara?")
|
47 |
+
```
|