Find answers from the community

Updated last year

Trying to use sub query question engine

Trying to use sub query question engine : I can see internally llamaindex is still using davinci. how can i change this to gpt3.5turbo?

Plain Text

\nGiven the new context, refine the original answer to better answer the question. If the context isn\'t useful, return the original answer.", "stream": false, "model": "text-davinci-003", "temperature": 0.0, "max_tokens": 1740}' message='Post details'

15 comments

kkeychron

here is my full code @Logan M @WhiteFang_Jr

LLogan M

You never passed the service context into the vector index 👀

kkeychron

i did

kkeychron

sub_query_engine = SubQuestionQueryEngine.from_defaults(
query_engine_tools=query_engine_tools,
service_context=service_context,
use_async=True,
)

kkeychron

i passed it into the subqery engine

kkeychron

can help me correct

kkeychron

this code is still working btw, just using davinci 3 internally

LLogan M

index= VectorStoreIndex.from_vector_store(vector_store = vector_store, storage_context=storage_context, service_context=service_context)

LLogan M

it was missing the service context on that line

WWhiteFang_Jr

Eagle eyes 😅

kkeychron

so service contect needs to go into query engine and vector index BOTH?

kkeychron

thanks

WWhiteFang_Jr

yes, or you can set global service context that will help you to not add it anywhere.

kkeychron

noted thanks

Add a reply