NotFoundError for OpenAI with gpt-3.5-turbo model

[shashankdeshpande]

Checked other resources

• I added a very descriptive title to this issue.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.
• I am sure that this is a bug in LangChain rather than my code.
• The bug is not resolved by updating to the latest stable version of LangChain (or the specific integration package).

Example Code

Getting error for this code -

from langchain_openai import OpenAI
llm = OpenAI(model='gpt-3.5-turbo', temperature=0, streaming=True)
llm('how are you?')

Error Message and Stack Trace (if applicable)

NotFoundError: Error code: 404 - {'error': {'message': 'This is a chat model and not supported in the v1/completions endpoint. Did you mean to use v1/chat/completions?', 'type': 'invalid_request_error', 'param': 'model', 'code': None}}

Description

Getting above specified error when configuring gpt-3.5-turbo model with OpenAI
However, this model works as expected with ChatOpenAI

On the other hand, gpt-3.5-turbo-instruct model also works as expected with OpenAI, code is mentioned below -

from langchain_openai import OpenAI
llm = OpenAI(model='gpt-3.5-turbo-instruct',temperature=0, streaming=True)
llm('how are you?')

Here is the screenshot for reference -

I believe this issue is due to configuring non-supported model with OpenAI instead of ChatOpenAI

Observation 🔍
I referred the codebase of openai python package and observed that there are some set of models which has only support of v1/chat/completions (ChatOpenAI as implemented within langchain). Check these files for more details -

• v1/chat/completions: src/openai/resources/chat/completions.py (list of specific models mentioned here)
• v1/completions: src/openai/resources/completions.py

Potential Fix 🤔
Should we validate the model name by referring to the same list when handling parameters for OpenAI, and raise an error accordingly?

I can work on this, please check and let me know

System Info

System Information

OS: Darwin
OS Version: Darwin Kernel Version 23.4.0: Fri Mar 15 00:11:05 PDT 2024; root:xnu-10063.101.17~1/RELEASE_X86_64
Python Version: 3.11.8 (main, Feb 26 2024, 15:43:17) [Clang 14.0.6 ]

Package Information

langchain_core: 0.1.42
langchain: 0.1.16
langchain_community: 0.0.32
langsmith: 0.1.47
langchain_openai: 0.1.3
langchain_text_splitters: 0.0.1
langchainhub: 0.1.15
openai: 1.17.0

Packages not installed (Not Necessarily a Problem)

The following packages were not found:

langgraph
langserve

[shashankdeshpande]

@hwchase17 @baskaryan @efriis Can you please check this?

[SzymonSt]

Bump I have the same problem latest langchain-openai==0.1.3

[baskaryan]

gpt-3.5 and gpt-4 are chat models so you must use from langchain_openai import ChatOpenAI. See https://python.langchain.com/docs/integrations/chat/openai/ for more

[shashankdeshpande]

@baskaryan Yeah, I got this. However, the problem is lack of documentation regarding which model serves as the chat model which might cause confusion. It might be beneficial to either clarify this in the documentation or implement an explicit error message. What are your thoughts on this?