Error "{'code': '404', 'message': 'Resource not found'}"

[mnbenavides]

Checked other resources

• I added a very descriptive title to this question.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.

Commit to Help

• I commit to help with one of those options 👆

Example Code

load_dotenv()

def get_embedding_function():

embeddings = AzureOpenAIEmbeddings(

    azure_endpoint= os.environ['AZURE_OPENAI_ENDPOINT'],

    deployment= "/deployments/text-embedding-ada-002/",

    openai_api_key= os.environ['AZURE_OPENAI_API_KEY'],

    openai_api_version = os.environ['OPENAI_API_VERSION'],

    model="text-embedding-ada-002",

)

return embeddings


from langchain_community.document_loaders import PyPDFDirectoryLoader

from langchain_text_splitters import RecursiveCharacterTextSplitter

from langchain.schema.document import Document

from get_embedding_function import get_embedding_function

from langchain_chroma import Chroma

def add_to_chroma(chunks: list[Document]):

# Load the existing database.

db = Chroma(

    collection_name="example_pdfs",

    persist_directory=CHROMA_PATH, 

    embedding_function=get_embedding_function()

)

chunks_with_ids = calculate_chunk_ids(chunks)

existing_items = db.get(include=[])

existing_ids = set(existing_items["ids"])

print(f"Number of existing documents in DB: {len(existing_ids)}")

# Only add documents that don't exist in the DB.

new_chunks = []

for chunk in chunks_with_ids:

    if chunk.metadata["id"] not in existing_ids:

        new_chunks.append(chunk)

if len(new_chunks):

    print(f"Adding new documents: {len(new_chunks)}")

    new_chunk_ids = [chunk.metadata["id"] for chunk in new_chunks]

    if len(new_chunks) == len(new_chunk_ids):

        try:

            db.add_documents(documents=new_chunks, ids=new_chunk_ids)

            #db.persist()

        except Exception as e:

            print(f"Error al agregar documentos a la base de datos: {e}")

    else:

        print(" Error: El número de documentos y el número de IDs no coincide.")

else:

    print("No new documents to add")

Description

I have already followed the steps provided:

• I double-checked the environment variables multiple times to ensure that AZURE_OPENAI_ENDPOINT and OPENAI_API_VERSION are correctly set.

• I confirmed that the deployment name exactly matches the "Model Deployment Name" in the Azure Portal.

• I installed the openai package and updated it to the latest version available.

*I tested the text-embedding-3-large, text-embedding-ada-002, and gpt-4o models to check for functionality.

• I also verified that the authentication with AZURE_OPENAI_API_KEY is correctly configured.

• Despite these efforts, I am still encountering the same issue. I would appreciate any additional assistance you can provide to help resolve this.

• log in to azure cli

Thank you in advance for your support.

System Info

"pip install -U langchain_openai"

Langchain (Version --> 0.2.14)
Phyton (Version --> 3.12.5)
Windows (Version --> Windows 11 pro )
Azure cli (Version --> 2.63.0)

[dosubot[bot]]

Hello @mnbenavides! I'm here to help you with any bugs, questions, or becoming a contributor. Let's squash those bugs together! 🤖

To resolve the "Resource not found" error when using AzureOpenAIEmbeddings in your Python code with LangChain, ensure that you have correctly set up and provided the necessary environment variables or parameters. Specifically, check the following:

1. Azure Endpoint: Ensure that the AZURE_OPENAI_ENDPOINT environment variable is correctly set to your Azure OpenAI endpoint URL. This should include the resource name and follow the format https://<your-endpoint>.openai.azure.com/.

2. API Key: Ensure that the AZURE_OPENAI_API_KEY environment variable is correctly set to your Azure OpenAI API key.

3. API Version: Ensure that the OPENAI_API_VERSION environment variable is correctly set to the appropriate API version, such as 2023-05-15.

4. Deployment Name: Ensure that you have specified the correct deployment name without any leading or trailing slashes. For example, use text-embedding-ada-002 instead of /deployments/text-embedding-ada-002/.

Here is an example of how to instantiate the AzureOpenAIEmbeddings class with the necessary parameters:

import os
from langchain_community.embeddings.openai import OpenAIEmbeddings

os.environ["AZURE_OPENAI_ENDPOINT"] = "https://<your-endpoint>.openai.azure.com/"
os.environ["AZURE_OPENAI_API_KEY"] = "your-api-key"
os.environ["OPENAI_API_VERSION"] = "2023-05-15"

def get_embedding_function():
    embeddings = OpenAIEmbeddings(
        azure_endpoint=os.environ['AZURE_OPENAI_ENDPOINT'],
        deployment="text-embedding-ada-002",
        openai_api_key=os.environ['AZURE_OPENAI_API_KEY'],
        openai_api_version=os.environ['OPENAI_API_VERSION'],
        model="text-embedding-ada-002",
    )
    return embeddings

Ensure that the environment variables are set correctly if you are not passing them directly:

export AZURE_OPENAI_API_KEY="your-api-key"
export AZURE_OPENAI_ENDPOINT="https://<your-endpoint>.openai.azure.com/"
export OPENAI_API_VERSION="2023-05-15"

By verifying and correctly setting these parameters, you should be able to resolve the "Resource not found" error ^[1][2].

To continue talking to Dosu, mention @dosu.

[mnbenavides]

Hi @dosu,

According to the new implementation, you used OpenAIEmbeddings, but according to the documentation, the correct model to implement is AzureOpenAIEmbeddings. You can review the details in the following link: AzureOpenAIEmbeddings Documentation.

I have verified the endpoints and attached evidence of this. Despite this, I am still receiving a 404 error. Could you suggest any other solution that I could implement to avoid this error?

Thank you for your help.

[dosubot[bot]]

To resolve the 404 "Resource not found" error when using AzureOpenAIEmbeddings, ensure the following:

1. Azure Endpoint: The azure_endpoint should be correctly set and should look like https://<your-endpoint>.openai.azure.com/.

2. Deployment Name: The deployment parameter should be set without any leading or trailing slashes. For example, use text-embedding-ada-002 instead of /deployments/text-embedding-ada-002/.

3. API Key: Ensure the openai_api_key is correctly set.

4. API Version: Ensure the openai_api_version is correctly set, such as 2023-05-15.

5. API Type: Ensure the openai_api_type is set to 'azure'.

Here is an updated example using AzureOpenAIEmbeddings:

import os
from langchain_openai import AzureOpenAIEmbeddings

os.environ["AZURE_OPENAI_ENDPOINT"] = "https://<your-endpoint>.openai.azure.com/"
os.environ["AZURE_OPENAI_API_KEY"] = "your-api-key"
os.environ["OPENAI_API_VERSION"] = "2023-05-15"

def get_embedding_function():
    embeddings = AzureOpenAIEmbeddings(
        azure_endpoint=os.environ['AZURE_OPENAI_ENDPOINT'],
        deployment="text-embedding-ada-002",
        openai_api_key=os.environ['AZURE_OPENAI_API_KEY'],
        openai_api_version=os.environ['OPENAI_API_VERSION'],
        openai_api_type='azure'
    )
    return embeddings

Ensure that the environment variables are set correctly if you are not passing them directly:

export AZURE_OPENAI_API_KEY="your-api-key"
export AZURE_OPENAI_ENDPOINT="https://<your-endpoint>.openai.azure.com/"
export OPENAI_API_VERSION="2023-05-15"

By verifying and correctly setting these parameters, you should be able to resolve the "Resource not found" error ^[1][2].

To continue talking to Dosu, mention @dosu.