force instantiation of OpenAIEmbeddings with openai instead of Azure

[wiswis15]

inside my project Im using asure open ai in some parts, but for a specific part, Im trying to use the langchain.embeddings.OpenAIEmbeddings class. with open ai and not azure.

when i do like this
from langchain.embeddings import OpenAIEmbeddings
embeddings = OpenAIEmbeddings.(openai_api_key=key)

it always instantiate it with azure, and then asks for the resource for example.

can anyone help please ?

[dosubot[bot]]

🤖

Hello,

From the context provided, it seems that the OpenAIEmbeddings class in the LangChain framework retrieves its configuration from environment variables. This means that whether it uses Azure or OpenAI depends on the values of these environment variables.

If you want to use OpenAI instead of Azure, you can set the OPENAI_API_TYPE environment variable to a value other than 'azure', 'azure_ad', or 'azuread'. You can do this in your Python code as follows:

import os

os.environ["OPENAI_API_TYPE"] = "openai"
os.environ["OPENAI_API_KEY"] = "your-openai-api-key"

Replace "your-openai-api-key" with your actual OpenAI API key.

After setting these environment variables, you can instantiate the OpenAIEmbeddings class as follows:

from langchain.embeddings import OpenAIEmbeddings

embeddings = OpenAIEmbeddings(openai_api_key=os.environ["OPENAI_API_KEY"])

This should instantiate the OpenAIEmbeddings class with OpenAI instead of Azure.

Please note that this solution is based on the current implementation of the OpenAIEmbeddings class in the LangChain codebase. If the implementation changes in the future, this solution may no longer work.

I hope this helps! If you have any further questions, please feel free to ask.

Best regards,
Dosu

Sources

• langchain.embeddings.OpenAIEmbeddings is not working with AzureOpenAI
• langchain AzureOpenAI Embeddings
• libs/langchain/langchain/embeddings/openai.py
• libs/langchain/langchain/llms/openai.py

---

This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.

Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.

If you want to continue the conversation, start your reply with @dosu-bot.

[wiswis15]

does not really work:
I have this exact code:
os.environ["OPENAI_API_TYPE"] = "openai"
os.environ["OPENAI_API_KEY"] = "blablabla"
embeddings = OpenAIEmbeddings(openai_api_key=os.environ["OPENAI_API_KEY"])
self.database = FAISS.from_texts(texts=all_splitted_text, embedding=embeddings,metadatas=metadata)

-->throws an exception:
File "/Users/wissem/workspace/TemujinApp/TemujinApp.git/temujin/temujin_core/temujin_openai/semantic_search_faiss.py", line 84, in create_db
self.database = FAISS.from_texts(texts=all_splitted_text, embedding=embeddings,metadatas=metadata)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/wissem/workspace/TemujinApp/TemujinApp.git/venv/lib/python3.11/site-packages/langchain/vectorstores/faiss.py", line 602, in from_texts
embeddings = embedding.embed_documents(texts)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/wissem/workspace/TemujinApp/TemujinApp.git/venv/lib/python3.11/site-packages/langchain/embeddings/openai.py", line 483, in embed_documents
return self._get_len_safe_embeddings(texts, engine=self.deployment)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/wissem/workspace/TemujinApp/TemujinApp.git/venv/lib/python3.11/site-packages/langchain/embeddings/openai.py", line 367, in _get_len_safe_embeddings
response = embed_with_retry(
^^^^^^^^^^^^^^^^^
File "/Users/wissem/workspace/TemujinApp/TemujinApp.git/venv/lib/python3.11/site-packages/langchain/embeddings/openai.py", line 107, in embed_with_retry
return _embed_with_retry(**kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/wissem/workspace/TemujinApp/TemujinApp.git/venv/lib/python3.11/site-packages/tenacity/init.py", line 289, in wrapped_f
return self(f, *args, **kw)
^^^^^^^^^^^^^^^^^^^^
File "/Users/wissem/workspace/TemujinApp/TemujinApp.git/venv/lib/python3.11/site-packages/tenacity/init.py", line 379, in call
do = self.iter(retry_state=retry_state)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/wissem/workspace/TemujinApp/TemujinApp.git/venv/lib/python3.11/site-packages/tenacity/init.py", line 314, in iter
return fut.result()
^^^^^^^^^^^^
File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/concurrent/futures/_base.py", line 449, in result
return self.__get_result()
^^^^^^^^^^^^^^^^^^^
File "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/concurrent/futures/_base.py", line 401, in __get_result
raise self._exception
File "/Users/wissem/workspace/TemujinApp/TemujinApp.git/venv/lib/python3.11/site-packages/tenacity/init.py", line 382, in call
result = fn(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^
File "/Users/wissem/workspace/TemujinApp/TemujinApp.git/venv/lib/python3.11/site-packages/langchain/embeddings/openai.py", line 104, in _embed_with_retry
response = embeddings.client.create(**kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/wissem/workspace/TemujinApp/TemujinApp.git/venv/lib/python3.11/site-packages/openai/api_resources/embedding.py", line 33, in create
response = super().create(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/wissem/workspace/TemujinApp/TemujinApp.git/venv/lib/python3.11/site-packages/openai/api_resources/abstract/engine_api_resource.py", line 153, in create
response, _, api_key = requestor.request(
^^^^^^^^^^^^^^^^^^
File "/Users/wissem/workspace/TemujinApp/TemujinApp.git/venv/lib/python3.11/site-packages/openai/api_requestor.py", line 298, in request
resp, got_stream = self._interpret_response(result, stream)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/wissem/workspace/TemujinApp/TemujinApp.git/venv/lib/python3.11/site-packages/openai/api_requestor.py", line 700, in _interpret_response
self._interpret_response_line(
File "/Users/wissem/workspace/TemujinApp/TemujinApp.git/venv/lib/python3.11/site-packages/openai/api_requestor.py", line 765, in _interpret_response_line
raise self.handle_error_response(
openai.error.InvalidRequestError: Resource not found

which mean it still instantiate it as azure one! or maybe the exception means something else ?

[mspronesti]

@wiswis15 Tried to reproduce your error using your snippet, to no avail. Can you provide a meaningful reproducible example which leads to the above error when you transition from Azure to "standard" OpenAI?

[wiswis15]

im using that same snippet inside a Django project where im already using Azure-openai for other parts, but i want to use normal openai for some other parts. that would be the difference between our environment I guess, it is really annoying.

[mspronesti]

Does passing openai_api_type="open_ai" to OpenAIEmbeddings happen to solve your problem?

embeddings = OpenAIEmbeddings(
        openai_api_key='sk-...' # OpenAI key, NOT Azure
        openai_api_type='open_ai'
) 

I never set it, unless I'm using Azure, but I'm just guessing you might have some place in your code where this value is set to azure and never changed.

[wiswis15]

i tried to set it just before the instantiation and same issue. Few people also mention hard time getting openai and azure openai to work within the same app.