Find answers from the community

Updated 2 years ago

yeah Yeah So I upgraded the that version

At a glance

yeah Yeah So I upgraded the that version but there is no GPTSimpleVectorIndex. and I am using GPTVectorStoreIndex instead GPTSimpleVectorIndex. but there is some issues also

10 comments

LLogan M

Check out the latest docs for the proper usage patterns and whatnot. A few things changed recently

https://gpt-index.readthedocs.io/en/latest/guides/primer/usage_pattern.html

ddirty_malik9k

Thank you

ddirty_malik9k

Could you let me know how to make the json file using GPTVectorStoreIndex?

Plain Text

index = GPTVectorStoreIndex.from_documents(documents)

    index.storage_context.persist(f"{data_directory}/index.json")

Is this right?

ddirty_malik9k

When use the GPTSimpleVectorIndex,
I used

Plain Text

index.save_to_disk('simple_index.json')

LLogan M

Yea, there's no more single file. Persist saves to a dir, usually 3 seperate files

LLogan M

So in the persist call, you should specify a folder

ddirty_malik9k

Ok and there is no GPTChromaIndex in current llama_index?

LLogan M

Correct, everything is unified to a single vector index, and now you just set the vector store underneath that index

Here's the chroma example
https://gpt-index.readthedocs.io/en/latest/examples/vector_stores/ChromaIndexDemo.html

ddirty_malik9k

Thank you

ddirty_malik9k

sorry for bothering you.
Could you let me know how to see the datas from ChromaDB collection?

Add a reply