From 37e84a403d6d11b670a42e84153204cd8b76b849 Mon Sep 17 00:00:00 2001
From: SangBin Cho <rkooo567@gmail.com>
Date: Tue, 16 Apr 2024 06:47:31 +0900
Subject: [PATCH] [Typing] Fix Sequence type GenericAlias only available after
 Python 3.9. (#4092)

---
 vllm/core/block_manager_v1.py | 5 +++--
 vllm/core/block_manager_v2.py | 2 +-
 vllm/core/interfaces.py       | 2 +-
 vllm/utils.py                 | 7 ++++---
 4 files changed, 9 insertions(+), 7 deletions(-)

diff --git a/vllm/core/block_manager_v1.py b/vllm/core/block_manager_v1.py
index e391a3b1e5a33..be093922b84f2 100644
--- a/vllm/core/block_manager_v1.py
+++ b/vllm/core/block_manager_v1.py
@@ -1,9 +1,10 @@
 """A block manager that manages token blocks."""
 from abc import ABC, abstractmethod
-from collections.abc import Sequence as GenericSequence
 from itertools import count, takewhile
 from os.path import commonprefix
-from typing import Dict, List, Optional, Set
+from typing import Dict, List, Optional
+from typing import Sequence as GenericSequence
+from typing import Set
 
 from vllm.block import BlockTable, PhysicalTokenBlock
 from vllm.core.evictor import EvictionPolicy, Evictor, make_evictor
diff --git a/vllm/core/block_manager_v2.py b/vllm/core/block_manager_v2.py
index 19f0cf415eb34..6339a6baf4161 100644
--- a/vllm/core/block_manager_v2.py
+++ b/vllm/core/block_manager_v2.py
@@ -1,6 +1,6 @@
 """A block manager that manages token blocks."""
-from collections.abc import Sequence as GenericSequence
 from typing import Dict, List, Optional
+from typing import Sequence as GenericSequence
 
 from vllm.core.block.block_table import BlockTable
 from vllm.core.block.cpu_gpu_block_allocator import CpuGpuBlockAllocator
diff --git a/vllm/core/interfaces.py b/vllm/core/interfaces.py
index c1f68a2e891bf..56c2c5995c38b 100644
--- a/vllm/core/interfaces.py
+++ b/vllm/core/interfaces.py
@@ -1,7 +1,7 @@
 import enum
 from abc import ABC, abstractmethod
-from collections.abc import Sequence as GenericSequence
 from typing import Dict, List
+from typing import Sequence as GenericSequence
 
 from vllm.sequence import Sequence, SequenceGroup
 
diff --git a/vllm/utils.py b/vllm/utils.py
index 4c0dc9ca729a9..aad62516ad1b9 100644
--- a/vllm/utils.py
+++ b/vllm/utils.py
@@ -6,11 +6,12 @@ import socket
 import subprocess
 import uuid
 import warnings
-from collections import OrderedDict, defaultdict
+from collections import defaultdict
 from functools import lru_cache, partial
 from platform import uname
 from typing import (Any, AsyncIterator, Awaitable, Callable, Dict, Generic,
-                    Hashable, List, Optional, Tuple, TypeVar, Union)
+                    Hashable, List, Optional, OrderedDict, Tuple, TypeVar,
+                    Union)
 
 import psutil
 import torch
@@ -51,7 +52,7 @@ class Counter:
 class LRUCache(Generic[T]):
 
     def __init__(self, capacity: int):
-        self.cache = OrderedDict[Hashable, T]()
+        self.cache: OrderedDict[Hashable, T] = OrderedDict()
         self.capacity = capacity
 
     def __contains__(self, key: Hashable) -> bool: