Update deprecated Python 3.8 typing (#13971)

vllm/v1/executor/abstract.py

@@ -1,7 +1,7 @@
 # SPDX-License-Identifier: Apache-2.0
 
 from concurrent.futures import Future
-from typing import List, Type, Union
+from typing import Union
 
 import torch
 import torch.distributed as dist
@@ -22,8 +22,8 @@ class Executor(ExecutorBase):
     For methods shared by v0 and v1, define them in ExecutorBase"""
 
     @staticmethod
-    def get_class(vllm_config: VllmConfig) -> Type["Executor"]:
-        executor_class: Type[Executor]
+    def get_class(vllm_config: VllmConfig) -> type["Executor"]:
+        executor_class: type[Executor]
         parallel_config = vllm_config.parallel_config
         distributed_executor_backend = (
             parallel_config.distributed_executor_backend)
@@ -53,7 +53,7 @@ class Executor(ExecutorBase):
         return executor_class
 
     def initialize_from_config(self,
-                               kv_cache_configs: List[KVCacheConfig]) -> None:
+                               kv_cache_configs: list[KVCacheConfig]) -> None:
         """
         Initialize the KV caches and begin the model execution loop of the
         underlying workers.
@@ -69,7 +69,7 @@ class Executor(ExecutorBase):
         # operators can be applied to all workers.
         return min(output)
 
-    def get_kv_cache_specs(self) -> List[KVCacheSpec]:
+    def get_kv_cache_specs(self) -> list[KVCacheSpec]:
         output = self.collective_rpc("get_kv_cache_spec")
         return output
 

vllm/v1/executor/multiproc_executor.py

@@ -10,7 +10,7 @@ from dataclasses import dataclass
 from enum import Enum, auto
 from functools import partial
 from multiprocessing.process import BaseProcess
-from typing import Any, Callable, Dict, List, Optional, Tuple, Union
+from typing import Any, Callable, Optional, Union
 
 import cloudpickle
 import psutil
@@ -77,7 +77,7 @@ class MultiprocExecutor(Executor):
         scheduler_output_handle = self.rpc_broadcast_mq.export_handle()
 
         # Create workers
-        self.workers: List[WorkerProcHandle] = []
+        self.workers: list[WorkerProcHandle] = []
         for rank in range(self.world_size):
             worker = WorkerProc.make_worker_process(self.vllm_config, rank,
                                                     rank,
@@ -94,8 +94,8 @@ class MultiprocExecutor(Executor):
     def collective_rpc(self,
                        method: Union[str, Callable],
                        timeout: Optional[float] = None,
-                       args: Tuple = (),
-                       kwargs: Optional[Dict] = None) -> List[Any]:
+                       args: tuple = (),
+                       kwargs: Optional[dict] = None) -> list[Any]:
         start_time = time.monotonic()
         kwargs = kwargs or {}
 
@@ -208,7 +208,7 @@ class WorkerProc:
         self.rank = rank
         wrapper = WorkerWrapperBase(vllm_config=vllm_config, rpc_rank=rank)
         # TODO: move `init_worker` to executor level as a collective rpc call
-        all_kwargs: List[Dict] = [
+        all_kwargs: list[dict] = [
             {} for _ in range(vllm_config.parallel_config.world_size)
         ]
         all_kwargs[rank] = {