I am using Langchain 0.2 to create a conversation pipeline with OpenAI LLM. Following is the code that I have written:

chat_llm = ChatOpenAI(
model_name="gpt-4",
api_key=config["api_key"],
openai_api_base=platform_url,
default_headers=llm_headers,
)

history_aware_retriever = create_history_aware_retriever(
llm=chat_llm,
retriever=vector_store.as_retriever(),
prompt=condense_default_system_question_prompt
)

qa_chain = create_stuff_documents_chain(chat_llm, qa_prompt)

convo_qa_chain = create_retrieval_chain(history_aware_retriever, qa_chain)

result = convo_qa_chain.invoke({
"input": user_query,
"chat_history": chat_memory.chat_memory.messages
})

enter image description here

On convo_qa_chain.invoke the OpenAI Api is giving the HTTP status 429(Too many requests).

NOTE: The Curl command with the same payload works. It gives me a valid output.

• I added logs.
• I extraced the payload and created a CURL Command that I found to be working.
• Debug but the abstraction of OpenAI extension provided in the Langchain_openai doesn’t let me understand any error.