I'm trying to use VectorStoreIndex
from llama_index
solving the RAG problem for chatbot just the following way:
import openai
from llama_index import VectorStoreIndex
index = VectorStoreIndex.from_documents(docs)
index.storage_context.persist()
When I read documentation, it is recommended to use openai.api_key = os.getenv('OPENAI_API_KEY')
to be able to connect model with OpenAI. But what should I do if I use AzureOpenAI (meaning, that I have api_key
, azure_endpoint
and api_version
and don't have openai api_key)? If I use openai.api_key = my_azure_api_key
this also does not work because apparently the model refers to https://platform.openai.com/account/api-keys
and not to AzureAI Services..
Sorry if the question looks dubious, was unable to find the information on web.
Thanks!
Have solved it. Actually all you need is to add context parameter into VectorStoreIndex.from_documents()
. In this parameter one can specify AzureOpenAI
or AzureOpenAIEmbedding
and obviously use api_key
, azure_endpoint
and api_version
.
index = VectorStoreIndex.from_documents(documents, service_context=service_context)