Method from_engine_args

TD_Pipe/engine/async_llm_engine.py:486–505

Creates an async LLM engine from the engine arguments.

(cls,
                         engine_args: AsyncEngineArgs,
                         start_engine_loop: bool = True)

    @classmethod
    def from_engine_args(cls,
                         engine_args: AsyncEngineArgs,
                         start_engine_loop: bool = True) -> "AsyncLLMEngine":
        """Creates an async LLM engine from the engine arguments."""
        # Create the engine configs.
        engine_configs = engine_args.create_engine_configs()
        parallel_config = engine_configs[2]
        # Initialize the cluster.
        placement_group = initialize_cluster(parallel_config,
                                             engine_args.engine_use_ray)
        # Create the async LLM engine.
        engine = cls(parallel_config.worker_use_ray,
                     engine_args.engine_use_ray,
                     *engine_configs,
                     placement_group,
                     log_requests=not engine_args.disable_log_requests,
                     log_stats=not engine_args.disable_log_stats,
                     max_log_len=engine_args.max_log_len,
                     start_engine_loop=start_engine_loop)
        return engine