[1/N] Initial prototype for multi-modal processor (#10044)

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>

vllm/engine/async_llm_engine.py

@@ -19,6 +19,7 @@ from vllm.executor.executor_base import ExecutorAsyncBase
 from vllm.executor.gpu_executor import GPUExecutorAsync
 from vllm.executor.ray_utils import initialize_ray_cluster
 from vllm.inputs import PromptType
+from vllm.inputs.preprocess import InputPreprocessor
 from vllm.logger import init_logger
 from vllm.lora.request import LoRARequest
 from vllm.model_executor.guided_decoding import (
@@ -729,6 +730,9 @@ class AsyncLLMEngine(EngineClient):
         self.set_errored(exc)
         self._request_tracker.propagate_exception(exc)
 
+    async def get_input_preprocessor(self) -> InputPreprocessor:
+        return self.engine.input_preprocessor
+
     async def get_tokenizer(
         self,
         lora_request: Optional[LoRARequest] = None,