Method record

utils/llm_client.py:65–69

(self, response: "LLMResponse")

        self._current_phase = phase

    def record(self, response: "LLMResponse") -> None:
        if self._current_phase == "memorize":
            self._memorize_usage.add(response.input_tokens, response.output_tokens, response.latency)
        else:
            self._query_usage.add(response.input_tokens, response.output_tokens, response.latency)

    def reset(self) -> None:
        self._memorize_usage = TokenUsage()