Not able to use GPT4 model with AzureOpenAI from from langchain.llms

[srithedesigner]

We used to use AzureOpenAI llm from langchain.llms with the text-davinci-003 model but after deploying GPT4 in Azure when trying to use GPT4, this error is being thrown:
Must provide an 'engine' or 'deployment_id' parameter to create a <class 'openai.api_resources.chat_completion.ChatCompletion'>

This is how we are initializing the model:

model = AzureOpenAI(
        streaming=streaming,
        client=openai.ChatCompletion(),
        callback_manager=callback,
        deployment_name= "gpt4",
        model_name="gpt-4-32k",
        openai_api_key=env.cloud.openai_api_key,
        temperature=temperature,
        max_tokens=max_tokens,
        verbose=verbose,
    )

How do we use GPT4 withe the AzureOpenAI chain? Is it currently being supported? or are we initializing it wrong?

[wYasha]

same issue

[GraviasAisera]

Same issue

[AshweeniKulkarni21]

I am facing the same issue

[mspronesti]

@srithedesigner have you tried langchain's AzureChatOpenAI? This works fine for me and I believe is the correct client for ChatCompletion.
Hope this helps.