Find answers from the community

Updated 2 months ago

Tools fit together

Hello everyone, this is not a technical question per se. But I am trying to build an internal tool for the startup that I work for. With the proliferation of LLM tools like Pinecone, Llama-index etc, its hard to see the forest for the trees. I guess my question is: How do the various tools fit with each other? For instance, how does Llama-index, a vector store like Pinecone, a tool builder like LangChain, and the LLM itself fir together.

I can move the question to another appropriate channel if the question is beyond the scope of discussion here. Thanks!!
L
b
L
7 comments
Basically, I see it like this:

  • pinecone (or similar like qdrant, chroma, etc) are good for storing lots of vectors/embeddings
  • llama index is used to connect your LLM app to your data, by building customizable and powerful indexes over your data. Use it when you need to build and persist a knowledge base.
  • langchain itself is more of a chat interface. It can use llama index as a "Tool", so that you can create and use your indexes from llama index for specific use cases inside langchain
  • the LLM is used in both langchain and llama index, to generate natural language for conversations and queries.
  • separately, you have an embedding model, which generates embeddings for documents that helps with search
Thank you for your message.

I think this makes things clearer. But I wonder: what is the distinction between a vector store and Llama index?

Oh is Llama index the "raw" data like the actual text whereas a vector store is a specialized tool for storing embeddings for the raw data ?
A vector store like a specialized place for putting vectors. If you are embedding a ton of documents, these vector stores are optimized for searching and storage

Llama index allows you to answer queries about that data, as well as building more complex data structures, all while leveraging the embeddings in vector stores.
That makes a lot of sense
I nominate this answer by @Logan M be turned into a post on the Llamaindex docs, along with diagram.
Might be a good idea! πŸ’‘
Thats a great idea.
I think it would go a long way in systematizing some of the awesome work that is being done by the open source community
Add a reply
Sign up and join the conversation on Discord