Find answers from the community

Updated 2 years ago

Storage

At a glance

Might be a dumb question, but not entirely sure:

Does using a vector_store persist the index automatically via the StorageContext? In my case, I'm using ChromaDB. Do I need to persist indexes to disk instead?

8 comments

LLogan M

Yea, it is automatic!

You can see the notebook here for a full guide
https://gpt-index.readthedocs.io/en/stable/examples/vector_stores/ChromaIndexDemo.html#basic-example-including-saving-to-disk

ccmagorian

Whats the right way to create an index from a vector_store instead of from documents?

ccmagorian

since in a separate process we loaded the documents and saved them automatically to the vector_store

LLogan M

yea, the notebook has an example of that 😉

Basically, setup the vector_store object to point to the existing vector store

Then, index = VectorStoreIndex.from_vector_store(vector_store)

ccmagorian

ah gotcha, and when we load documents into the vector_store its automatically loaded?

LLogan M

sorry, what did you mean by that?

ccmagorian

automatically saved, my apologies

LLogan M

Yea it is automatically saved when you load documents 👍

Add a reply