LlamaIndex

Log inLog into community

Find answers from the community

Updated 7 months ago

Hi!

Hi!

At a glance

The community member is using LLAMA with Azure OpenAI to build a natural language to SQL conversion system. They are facing issues with configuring the VectorIndex and getting an authentication error. The community members suggest using the correct API key, version, and deployment name, as well as ensuring the LLM and embedding models are properly configured. They try using Azure OpenAI for both the LLM and embedding models, but still face issues with the embedding model. After several attempts, the community member is able to resolve the issue, but the summary does not explicitly state the solution.

Useful resources

·

Hi!
We are using LLAMA with Azure open ai, we are building the NL to SQL using the sql connect using sqlalchemy createnegine, trying to configure VectorIndex using the below code, while running the below code I am getting openai.AuthenticationError
obj_index = ObjectIndex.from_objects(
table_schema_objs,
table_node_mapping,
VectorStoreIndex,
# llm=llm
)

Error:
openai.AuthenticationError: Error code: 401 - {'error': {'message': 'Incorrect API key provided: 521fdde6**4b04. You can find your API key at https://platform.openai.com/account/api-keys.', 'type': 'invalid_request_error', 'param': None, 'code': 'invalid_api_key'}}

I am using the correct API key, version and deployment name, I have checked it.

Please suggest what is wrong with it.

W

U

37 comments

You have commented down your llm instance in the given above code.

Is this intentional ?

Try using Settings

Plain Text

from llama_index.core import Settings
Settings.llm = llm # your azure instance
Settings.embed_model = embed_model # your embedding mode instance

Hi @WhiteFang_Jr ,
Yes, it was intentional, I tried adding above code, but same response.
llm = AzureOpenAI(
engine="controltower", model="gpt-35-turbo", temperature=0.0
)
embed_model=OpenAIEmbedding(model_name="gpt-35-turbo")
Settings.llm = llm # your azure instance
Settings.embed_model = embed_model # your embedding mode instance

obj_index = ObjectIndex.from_objects(
table_schema_objs,
table_node_mapping,
VectorStoreIndex,
llm=llm
)

Ah I see the problem.
You have used azure openai for llm and normal openai for embedding

Plus you have defined wrong name in embedding model.

Either use full azure or full openai

@WhiteFang_Jr I tried using complete Azure.
import os
import openai
from llama_index.core import SQLDatabase
from openai import AzureOpenAI
from llama_index.llms.azure_openai import AzureOpenAI
from langchain_openai import AzureOpenAIEmbeddings

llm = AzureOpenAI(
engine="controltower", model="gpt-35-turbo", temperature=0.0
)

embed_model=AzureOpenAIEmbeddings(model="gpt-35-turbo", azure_endpoint="https://ccllmopenai.openai.azure.com/")

Settings.llm = llm # your azure instance
Settings.embed_model = embed_model # your embedding mode instance

obj_index = ObjectIndex.from_objects(
table_schema_objs,
table_node_mapping,
VectorStoreIndex,
llm=llm
)

Still facing issues, I tried model with gpt-4o, gpt-35-turbo, gpt-4.
can you please suggest what should be the model for NLtoSQL.

You are still putting an incorrect model name in embedding.

Do the following and try once

pip install llama-index-embeddings-azure-openai

Then import it like this
from llama_index.embeddings.azure_openai import AzureOpenAIEmbedding

Settings.embed_model = AzureOpenAIEmbedding()

In the embedding model you are passing llm model name which is why you might be getting the error.

now I am getting below error:
Retrying llama_index.embeddings.openai.base.get_embeddings in 3.913704518450162 seconds as it raised APIConnectionError: Connection error..

raise UnsupportedProtocol(
httpcore.UnsupportedProtocol: Request URL is missing an 'http://' or 'https://' protocol.

The above exception was the direct cause of the following exception:

File "/Users/kiran-rayabandi/.pyenv/versions/3.9.2/lib/python3.9/site-packages/openai/_base_client.py", line 986, in _request
raise APIConnectionError(request=request) from err
openai.APIConnectionError: Connection error.

Try providing the azure_endpoint: https://github.com/run-llama/llama_index/blob/7b14500fa41a37faff764742741cca4c6406e580/llama-index-integrations/embeddings/llama-index-embeddings-azure-openai/llama_index/embeddings/azure_openai/base.py#L18

Provide details like the given below doc:https://docs.llamaindex.ai/en/stable/examples/docstore/AzureDocstoreDemo/?h=azure#define-our-models

Do add azure endpoint

I changed the embedding model to text-embedding-ada-002, it worked.
but I am connecting to SQL, to process natural language to SQL

obj_index = ObjectIndex.from_objects(
table_schema_objs,
table_node_mapping,
VectorStoreIndex,
llm=llm
)
query_engine = SQLTableRetrieverQueryEngine(
sql_database, obj_index.as_retriever(similarity_top_k=1)
)

print(query_engine)

response = query_engine.query("Show me the inventory")

response.metadata["result"]

facing below error
ValueError: Unknown model 'text-embedding-ada-002'. Please provide a valid OpenAI model name in: gpt-4, gpt-4-32k, gpt-4-1106-preview, gpt-4-0125-preview, gpt-4-turbo-preview, gpt-4-vision-preview, gpt-4-1106-vision-preview, gpt-4-turbo-2024-04-09, gpt-4-turbo, gpt-4o, gpt-4o-2024-05-13, gpt-4-0613, gpt-4-32k-0613, gpt-4-0314, gpt-4-32k-0314, gpt-3.5-turbo, gpt-3.5-turbo-16k, gpt-3.5-turbo-0125, gpt-3.5-turbo-1106, gpt-3.5-turbo-0613, gpt-3.5-turbo-16k-0613, gpt-3.5-turbo-0301, text-davinci-003, text-davinci-002, gpt-3.5-turbo-instruct, text-ada-001, text-babbage-001, text-curie-001, ada, babbage, curie, davinci, gpt-35-turbo-16k, gpt-35-turbo, gpt-35-turbo-0125, gpt-35-turbo-1106, gpt-35-turbo-0613, gpt-35-turbo-16k-0613

with embedded model, query is not working

If you look at the error, you are choosing wrong model name

yes, but if I choose gpt-35-turbo-16k,gpt-4 or any other getting below error
ValueError: 'gpt-35-turbo-16k' is not a valid OpenAIEmbeddingModelType

The model that you have provided is not the valid model name.

You can find the valid model name in this: https://github.com/run-llama/llama_index/blob/7b14500fa41a37faff764742741cca4c6406e580/llama-index-integrations/embeddings/llama-index-embeddings-openai/llama_index/embeddings/openai/base.py#L35

if it is NL to SQL should we not be using the embedded model?

These model names are for llm

They are not for embedding purpose

Error I got using text-embedding-ada-002
ValueError: Unknown model 'text-embedding-ada-002'. Please provide a valid OpenAI model name in: gpt-4, gpt-4-32k, gpt-4-1106-preview, gpt-4-0125-preview, gpt-4-turbo-preview, gpt-4-vision-preview, gpt-4-1106-vision-preview, gpt-4-turbo-2024-04-09, gpt-4-turbo, gpt-4o, gpt-4o-2024-05-13, gpt-4-0613, gpt-4-32k-0613, gpt-4-0314, gpt-4-32k-0314, gpt-3.5-turbo, gpt-3.5-turbo-16k, gpt-3.5-turbo-0125, gpt-3.5-turbo-1106, gpt-3.5-turbo-0613, gpt-3.5-turbo-16k-0613, gpt-3.5-turbo-0301, text-davinci-003, text-davinci-002, gpt-3.5-turbo-instruct, text-ada-001, text-babbage-001, text-curie-001, ada, babbage, curie, davinci, gpt-35-turbo-16k, gpt-35-turbo, gpt-35-turbo-0125, gpt-35-turbo-1106, gpt-35-turbo-0613, gpt-35-turbo-16k-0613

These are options it accepst

_QUERY_MODE_MODEL_DICT = {
(OAEM.SIMILARITY_MODE, "davinci"): OAEMM.TEXT_SIMILARITY_DAVINCI,
(OAEM.SIMILARITY_MODE, "curie"): OAEMM.TEXT_SIMILARITY_CURIE,
(OAEM.SIMILARITY_MODE, "babbage"): OAEMM.TEXT_SIMILARITY_BABBAGE,
(OAEM.SIMILARITY_MODE, "ada"): OAEMM.TEXT_SIMILARITY_ADA,
(OAEM.SIMILARITY_MODE, "text-embedding-ada-002"): OAEMM.TEXT_EMBED_ADA_002,
(OAEM.SIMILARITY_MODE, "text-embedding-3-small"): OAEMM.TEXT_EMBED_3_SMALL,
(OAEM.SIMILARITY_MODE, "text-embedding-3-large"): OAEMM.TEXT_EMBED_3_LARGE,
(OAEM.TEXT_SEARCH_MODE, "davinci"): OAEMM.TEXT_SEARCH_DAVINCI_QUERY,
(OAEM.TEXT_SEARCH_MODE, "curie"): OAEMM.TEXT_SEARCH_CURIE_QUERY,
(OAEM.TEXT_SEARCH_MODE, "babbage"): OAEMM.TEXT_SEARCH_BABBAGE_QUERY,
(OAEM.TEXT_SEARCH_MODE, "ada"): OAEMM.TEXT_SEARCH_ADA_QUERY,
(OAEM.TEXT_SEARCH_MODE, "text-embedding-ada-002"): OAEMM.TEXT_EMBED_ADA_002,
(OAEM.TEXT_SEARCH_MODE, "text-embedding-3-large"): OAEMM.TEXT_EMBED_3_LARGE,
(OAEM.TEXT_SEARCH_MODE, "text-embedding-3-small"): OAEMM.TEXT_EMBED_3_SMALL,
}

but I am unable to craete them on the Azure Open AI

I am not sure hot to handle it

Please let me if you can connect on a call?

You'll have to create an instance of embedding model on azure first

Is that step done?

Attachment

I did created deployments in the azure open ai

These are llms, you'll have to add a embedding model

I might have not created it

https://github.com/openai/openai-cookbook/blob/main/examples%2Fazure%2Fembeddings.ipynb

This page has some info on how to deploy embedding model

You can refer to this

Thanks!

let me do that and update you

another error
openai.BadRequestError: Error code: 400 - {'error': {'code': 'OperationNotSupported', 'message': 'The embeddings operation does not work with the specified model, gpt-35-turbo-16k. Please choose different model and try again. You can learn more about which models can be used with each operation here: https://go.microsoft.com/fwlink/?linkid=2197993.'}}

Can you share your full code as I think you are still putting the wrong model name for embedding

please check the code

HI @WhiteFang_Jr
I am able to fix it

thanks for helping me out

Add a reply

Sign up and join the conversation on Discord