A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 17, 00:30 UTC
[Bugfix] Remove deprecated @abstractproperty (#5174)
Zhuohan Li committed
Jun 1, 2024 at 15:40 UTC
8279078e218833b357f7c5076850e3688714d570
3 files changed
+9
-6
vllm/core/evictor_v1.py
+3
-2
@@ -1,5 +1,5 @@
1
import enum
2
-from abc import ABC, abstractmethod, abstractproperty
2
+from abc import ABC, abstractmethod
3
from typing import OrderedDict
4
5
from vllm.block import PhysicalTokenBlock
@@ -44,7 +44,8 @@ class Evictor(ABC):
44
"""
45
pass
46
47
- @abstractproperty
47
+ @property
48
+ @abstractmethod
49
def num_blocks(self) -> int:
50
pass
51
vllm/core/evictor_v2.py
+3
-2
@@ -1,5 +1,5 @@
1
import enum
2
-from abc import ABC, abstractmethod, abstractproperty
2
+from abc import ABC, abstractmethod
3
from typing import OrderedDict, Tuple
4
5
@@ -46,7 +46,8 @@ class Evictor(ABC):
46
"""Remove a given block id from the cache."""
47
pass
48
49
- @abstractproperty
49
+ @property
50
+ @abstractmethod
51
def num_blocks(self) -> int:
52
pass
53
vllm/lora/worker_manager.py
+3
-2
@@ -1,4 +1,4 @@
1
-from abc import ABC, abstractmethod, abstractproperty
1
+from abc import ABC, abstractmethod
2
from contextlib import contextmanager
3
from typing import Any, Dict, List, Literal, Optional, Set, Type, Union
4
@@ -42,7 +42,8 @@ class AbstractWorkerLoRAManager(ABC):
42
yield
43
self._cached_dummy_lora = False
44
45
- @abstractproperty
45
+ @property
46
+ @abstractmethod
47
def is_enabled(self) -> bool:
48
...
49