VLLM代码解读 | VLLM Hack 3

在上一期，我们看到了多个输入如何被封装，然后被塞入llm_engine中，接下来，通过_run_engine,我们要进行输入的处理了。

    def _run_engine(
            self, *, use_tqdm: bool
    ) -> List[Union[RequestOutput, EmbeddingRequestOutput]]:
        # Initialize tqdm.
        if use_tqdm:
            num_requests = self.llm_engine.get_num_unfinished_requests()
            pbar = tqdm(
                total=num_requests,
                desc="Processed prompts",
                dynamic_ncols=True,
                postfix=f"Generation Speed: {0:.2f} toks/s",
            )
        # Run the engine.
        outputs: List[Union[RequestOutput, EmbeddingRequestOutput]] = []
        total_toks = 0
        while self.llm_engine.has_unfinished_requests():
            step_outputs = self.llm_engine.step()
            for output in step_outputs:
                if output.finished:
                    outputs.append(output)
                    if use_tqdm:
                        if isinstance(output, RequestOutput):
                            # Calculate tokens only for RequestOutput
                            total_toks += sum(
                                len(stp.token_ids) for stp in output.outputs)
                            spd = total_toks / pbar.format_dict["elapsed"]
                            pbar.postfix = f"Generation Speed: {spd:.2f} toks/s"
                        pbar.update(1)
        if use_tqdm:
            pbar.close()
        # Sort the outputs by request ID.
        # This is necessary