Hector RAG

A modular and extensible RAG (Retrieval Augmented Generation) package built on PostgreSQL vector database, offering advanced retrieval methods and fusion capabilities.

Key Features

• Multiple RAG retrieval methods:
  • Similarity Search
  • Keyword Search
  • Graph Retrieval
  • KAG (Knowledge-Aware Graph)
• Reciprocal Rank Fusion (RRF) for combining multiple retrieval methods
• Built on PostgreSQL vector database for efficient vector storage and retrieval
• Modular architecture allowing easy integration and customization
• Advanced RAG pipeline creation capabilities

Installation

Using pip:

pip install hector_rag

Using Poetry:

poetry add hector_rag

Requirements

• Python >=3.10,<3.13
• PostgreSQL database
• Dependencies:
  • networkx
  • semantic-router
  • pgvector
  • sqlalchemy

Quick Start

Basic Usage - Using 1 pg retriever

import os

from hector_rag import Hector
from hector_rag.retrievers import SimilarityRetriever, KeywordRetriever, GraphRetriever, RRFHybridRetriever
from hector_rag import Hector
from hector_rag.retrievers import GraphRetriever, SemanticRetriever, KeywordRetriever

semantic_retriever = SemanticRetriever(cursor,embeddings,embeddings_dimension=1536,collection_name=collection_name)
semantic_retriever.init_tables()
resp = semantic_retriever.get_relevant_documents(query="What is Fetch Ai ?", document_limit=10)

print(resp)

Advanced Usage

Combining Multiple Retrievers with RRF

import os

from hector_rag import Hector
from hector_rag.retrievers import SimilarityRetriever, KeywordRetriever, GraphRetriever, RRFHybridRetriever
from hector_rag import Hector
from hector_rag.retrievers import GraphRetriever, SemanticRetriever, KeywordRetriever

from langchain_openai.embeddings import OpenAIEmbeddings
from langchain_openai import ChatOpenAI

os.environ["OPENAI_API_KEY"] = os.getenv("OPENAI_API_KEY")

llm = ChatOpenAI(model="gpt-3.5-turbo")
embeddings = OpenAIEmbeddings(model="text-embedding-3-small")
collection_name = "new_collection_1"

rag = Hector(connection,embeddings, collection_name, {})

# Init all the retrievers that you want to use
semantic_retriever = SemanticRetriever()
graph_retriever = GraphRetriever(llm=llm)
keyword_retriever = KeywordRetriever()

# Add retrievers to Rag pipeline
rag.add_retriever(semantic_retriever)
rag.add_retriever(graph_retriever)
rag.add_retriever(keyword_retriever)

# Fetch documents
docs = rag.get_relevant_documents("What is  Decentralized AI ?", document_limit=10)

# Or directly use Hector Invoke to get llm response

while True:
    query = str(input("Enter query: "))
    response = rag.invoke(llm,query)
    print(response)

Development

To set up the development environment:

# Clone the repository
git clone https://github.com/yourusername/hector-rag.git
cd hector-rag

# Install dependencies using Poetry
poetry install

Testing

poetry run pytest

Documentation

For detailed documentation about each retriever type and fusion methods, please visit our documentation page.

Contributing

Contributions are welcome! Whether it's:

• Adding new retrieval methods
• Improving existing retrievers
• Enhancing documentation
• Reporting bugs
• Suggesting features

Please feel free to submit a Pull Request or create an Issue.

License

MIT License

Contact

For issues and feature requests, please use the GitHub Issues page.

Would you like me to add or modify any specific section of this README?