Method chat_stream

web/pgadmin/llm/providers/docker.py:380–419

Stream a chat response from Docker Model Runner.

(
        self,
        messages: list[Message],
        tools: Optional[list[Tool]] = None,
        system_prompt: Optional[str] = None,
        max_tokens: int = 4096,
        temperature: float = 0.0,
        **kwargs
    )

        )

    def chat_stream(
        self,
        messages: list[Message],
        tools: Optional[list[Tool]] = None,
        system_prompt: Optional[str] = None,
        max_tokens: int = 4096,
        temperature: float = 0.0,
        **kwargs
    ) -> Generator[Union[str, LLMResponse], None, None]:
        """Stream a chat response from Docker Model Runner."""
        converted_messages = self._convert_messages(messages)

        if system_prompt:
            converted_messages.insert(0, {
                'role': 'system',
                'content': system_prompt
            })

        payload = {
            'model': self._model,
            'messages': converted_messages,
            'max_completion_tokens': max_tokens,
            'temperature': temperature,
            'stream': True,
            'stream_options': {'include_usage': True}
        }

        if tools:
            payload['tools'] = self._convert_tools(tools)
            payload['tool_choice'] = 'auto'

        try:
            yield from self._process_stream(payload)
        except LLMClientError:
            raise
        except Exception as e:
            raise LLMClientError(LLMError(
                message=f"Streaming request failed: {str(e)}",
                provider=self.provider_name
            ))

    def _process_stream(
        self, payload: dict