Method model_dump

optillm/inference.py:1712–1741

(self)

        self.usage = ChatCompletionUsage(**response_dict["usage"])
    
    def model_dump(self) -> Dict:
        return {
            "id": self.id,
            "object": self.object,
            "created": self.created,
            "model": self.model,
            "choices": [
                {
                    "index": choice.index,
                    "message": {
                        "role": choice.message.role,
                        "content": choice.message.content,
                        "logprobs": choice.message.logprobs
                    } if choice.message.logprobs else {
                        "role": choice.message.role,
                        "content": choice.message.content
                    },
                    "finish_reason": choice.finish_reason
                }
                for choice in self.choices
            ],
            "usage": {
                "prompt_tokens": self.usage.prompt_tokens,
                "completion_tokens": self.usage.completion_tokens,
                "total_tokens": self.usage.total_tokens,
                "completion_tokens_details": {
                    "reasoning_tokens": getattr(self.usage, 'reasoning_tokens', 0)
                }
            }
        }

class InferenceClient:
    """OpenAI SDK Compatible client for local inference with dynamic model support"""