hi this is another beginner question is

At a glance

hi this is another beginner question is there anyway to call upon an existing collection rather then recreating and reindexing it

31 comments

LLogan M

When doing from_vector_strore, you don't need to pass in the storage context again or set it up, just pass in the vector store and service context

aaszaiman1

index = VectorStoreIndex.from_vector_store(
vector_store=vector_store,
service_context=service_context,
)" ----------- like this?

LLogan M

I think so!

ssylvinbestin

Hi @Logan M , I have been trying to load from vector, using the same method, and it doesn't seem to work

ssylvinbestin

client = weaviate.Client("http://localhost:8080")
self.vector_store = WeaviateVectorStore(weaviate_client=client, class_name='LlamaIndex_solution')
self._index = VectorStoreIndex.from_vector_store(vector_store=self.vector_store)

ssylvinbestin

that index doesn't find anything when i try to search

ssylvinbestin

i'm able to use a weaviatereader to read the documents from the same class in weaviate

ssylvinbestin

and build the index using from_documents

ssylvinbestin

of course that defeats the purpose of using a vector store

LLogan M

I can spin up weaviate and test, but that looks correct at first glance

ssylvinbestin

been pulling my hair out trying to figure out what i'm doing wrong

ssylvinbestin

yeah would very much appreciate it

ssylvinbestin

the data is definitely in weaviate in my instance

ssylvinbestin

it seems like it's trying to create a new index

ssylvinbestin

weaviate gets a new class each time

ssylvinbestin

{"action":"hnsw_vector_cache_prefill","count":1000,"index_id":"llamaindex_4cbf2e5574dc4e6689902c06216916d5_2XomdgHy9FiZ","level":"info","limit":1000000000000,"msg":"prefilled vector cache","time":"2023-07-21T22:31:52Z","took":124200}

LLogan M

ive 100% loaded an index from weaviate before, just need to refresh myself with how it works again

ssylvinbestin

thank you

ssylvinbestin

i must be missing smething obvious

ssylvinbestin

but can't find it even tracing through code

ssylvinbestin

at least so far

LLogan M

hmm tbh it worked for me lol

One misleading thing is it says class_prefix is deprecated, please use class_name but actually, you should use index_name lol

Attachment

LLogan M

Here's all my code

Plain Text

client = weaviate.Client("https://llama-test-8dkemtip.weaviate.network", auth_client_secret=resource_owner_config)
loaded_vector_store = WeaviateVectorStore(weaviate_client=client, class_prefix="LlamaIndex_solution")
loaded_index = VectorStoreIndex.from_vector_store(loaded_vector_store)
query_engine = index.as_query_engine()
response = query_engine.query("What happened at interleaf?")
print(response)

that's it!

index name

that's what got me

works now

i'll do a pr for the doc

ssylvinbestin

thank you so mich