list index out of range error using DeepSeek

[ThomasCosyn]

Hello,

I tried to integrate a DeepSeek model deployed on Azure with AzureAIChatCompletionsModel, but I get the error

File site-packages\langchain_azure_ai\chat_models\inference.py", line 500, in _stream  
choice = chunk.choices[0]  
IndexError : list index out of range

I guess the API sometimes sends a response with a choices empty list. We could check the size of the list before getting the first item ?

        self,
        messages: List[BaseMessage],
        stop: Optional[List[str]] = None,
        run_manager: Optional[CallbackManagerForLLMRun] = None,
        **kwargs: Any,
    ) -> Iterator[ChatGenerationChunk]:
        inference_messages = to_inference_message(messages)
        default_chunk_class = AIMessageChunk
        response = self._client.complete(
            messages=inference_messages,
            stream=True,
            stop=stop or self.stop,
            **self._identifying_params,
            **kwargs,
        )
        for chunk in response:
            if len(chunk.choices) > 0:
                choice = chunk.choices[0]
                chunk = _convert_delta_to_message_chunk(choice.delta, default_chunk_class)
                finish_reason = choice.finish_reason
                generation_info = (
                    dict(finish_reason=finish_reason) if finish_reason is not None else None
                )
                default_chunk_class = chunk.__class__  # type: ignore[assignment]
                cg_chunk = ChatGenerationChunk(
                    message=chunk, generation_info=generation_info
                )
                if run_manager:
                    run_manager.on_llm_new_token(cg_chunk.text, chunk=cg_chunk)
                yield cg_chunk```