Pinecone

nnyx-12

Hey guys, I'm trying to index documents loaded in pinecone as below. However I get this error. Anyone with an idea on how I can solve this?

AttributeError: 'str' object has no attribute 'upsert'
----> 6 index = GPTVectorStoreIndex.from_documents(documents, storage_context=storage_context)

Code chunk:
coppervstore = PineconeVectorStore(
pinecone_index=index_name,
add_sparse_vector=True,
)
storage_context = StorageContext.from_defaults(vector_store=coppervstore)
index = GPTVectorStoreIndex.from_documents(documents, storage_context=storage_context)

14 comments

LLogan M

I think that pinecone_index is more than a string

LLogan M

See this example

LLogan M

Plain Text

api_key = "api_key"
pinecone.init(api_key=api_key, environment="us-west1-gcp")
pinecone.create_index(
    "quickstart",
    dimension=1536,
    metric="euclidean",
    pod_type="p1"
)
pinecone_index = pinecone.Index("quickstart")

LLogan M

Then that pinecone index could go into the vector store.

Once the index is created, you can of course skip that line that creates it

nnyx-12

@Logan M Oooh makes sense, will try this. Thank you.

nnyx-12

But actually the main problem is in this part
index = GPTVectorStoreIndex.from_documents(documents, storage_context=storage_context)

nnyx-12

So using GPTVectorStoreIndex results to that error. I already created the pinecone_index

LLogan M

What error does that give you? That shouuuuld work

LLogan M

https://gpt-index.readthedocs.io/en/latest/examples/vector_stores/PineconeIndexDemo.html

LLogan M

Also side note when you get there, If you've already inserted the documents with llama index, you can just do this to reconnect to it GPTVectorStoreIndex([], storage_context=storage_context)

nnyx-12

Let me try this.. The error is

AttributeError Traceback (most recent call last)
<ipython-input-27-bddaeb9f6a4b> in <cell line: 1>()
----> 1 index = GPTVectorStoreIndex.from_documents(documents, storage_context=storage_context)

7 frames
/usr/local/lib/python3.10/dist-packages/llama_index/vector_stores/pinecone.py in add(self, embedding_results)
207 )[0]
208 entry.update({"sparse_values": sparse_vector})
--> 209 self._pinecone_index.upsert(
210 [entry], namespace=self._namespace, **self._insert_kwargs
211 )

AttributeError: 'str' object has no attribute 'upsert'

LLogan M

The pinecone index is still a string 👀

LLogan M

You set it up like the code example I gave?

nnyx-12

Its now working, yes I followed the instructions. Thank you for your help.

Add a reply

Find answers from the community

Pinecone