Pinecone Slim

PineconeSlim Class Documentation

Overview

The PineconeSlim class is a slim vector store of Pinecone, optimized for RAGchain.

Usage

Its usage is similar to Langchain's Pinecone class. Due to its inheritance to Pinecone class, you can use any function in Pinecone class. Plus, you can initialize with Pinecone class initializer.

For the optimal utilization of PineconeSlim VectorStore, developers are strongly recommended to employ the add_passages() method when saving passages. The use of alternative methods may not fully leverage the capabilities and advantages inherent in Slim VectorStore.

Here is a basic example:

import pinecone
from RAGchain.utils.vectorstore import PineconeSlim
from RAGchain.utils.embed import EmbeddingFactory

index = pinecone.Index(index_name="your_index_name")

# Initialize PineconeSlim
pinecone_slim = PineconeSlim(
    index=index,
    embedding=EmbeddingFactory('openai').get(),
    text_key="text",
    namespace="your_namespace"
)

# add passage to PineconeSlim
passages =[...list_of_passages...] # Assume we have list_of_passages retrieved earlier
pinecone_slim.add_passages(passages)