Update Optional[x] -> x | None and Union[x, y] to x | y (#26633)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
@@ -2,9 +2,8 @@
|
||||
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
|
||||
|
||||
import weakref
|
||||
from collections.abc import Sequence
|
||||
from collections.abc import Callable, Sequence
|
||||
from copy import deepcopy
|
||||
from typing import Callable, Union
|
||||
|
||||
from torch import fx
|
||||
from torch._ops import OpOverload
|
||||
@@ -44,7 +43,7 @@ class TestBackend:
|
||||
Inductor config is default-initialized from VllmConfig.CompilationConfig.
|
||||
"""
|
||||
|
||||
def __init__(self, *passes: Union[InductorPass, Callable[[fx.Graph], None]]):
|
||||
def __init__(self, *passes: InductorPass | Callable[[fx.Graph], None]):
|
||||
self.custom_passes = list(passes)
|
||||
compile_config = get_current_vllm_config().compilation_config
|
||||
self.inductor_config = compile_config.inductor_compile_config
|
||||
|
||||
@@ -10,7 +10,7 @@ initialized randomly with a fixed seed.
|
||||
"""
|
||||
|
||||
from dataclasses import dataclass
|
||||
from typing import Any, Optional
|
||||
from typing import Any
|
||||
|
||||
import pytest
|
||||
import torch
|
||||
@@ -162,7 +162,7 @@ class LlamaDecoderLayer(nn.Module):
|
||||
self,
|
||||
positions: torch.Tensor,
|
||||
hidden_states: torch.Tensor,
|
||||
residual: Optional[torch.Tensor],
|
||||
residual: torch.Tensor | None,
|
||||
) -> tuple[torch.Tensor, torch.Tensor]:
|
||||
"""
|
||||
For tractable computation:
|
||||
@@ -217,7 +217,7 @@ class LlamaModel(nn.Module):
|
||||
|
||||
def forward(
|
||||
self,
|
||||
input_ids: Optional[torch.Tensor],
|
||||
input_ids: torch.Tensor | None,
|
||||
positions: torch.Tensor,
|
||||
) -> torch.Tensor:
|
||||
hidden_states = self.embedding_tokens(input_ids)
|
||||
|
||||
@@ -1,7 +1,5 @@
|
||||
# SPDX-License-Identifier: Apache-2.0
|
||||
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
|
||||
from __future__ import annotations
|
||||
|
||||
import dataclasses
|
||||
|
||||
import pytest
|
||||
|
||||
@@ -1,11 +1,9 @@
|
||||
# SPDX-License-Identifier: Apache-2.0
|
||||
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
import logging
|
||||
import tempfile
|
||||
from typing import Any, Union
|
||||
from typing import Any
|
||||
|
||||
import pytest
|
||||
import torch
|
||||
@@ -217,7 +215,7 @@ def test_inductor_graph_partition_attn_fusion(caplog_vllm):
|
||||
|
||||
|
||||
def run_model(
|
||||
compile_config: Union[int, CompilationConfig],
|
||||
compile_config: int | CompilationConfig,
|
||||
model: str,
|
||||
model_kwargs: dict[str, Any],
|
||||
):
|
||||
|
||||
@@ -1,7 +1,6 @@
|
||||
# SPDX-License-Identifier: Apache-2.0
|
||||
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
|
||||
import copy
|
||||
from typing import Optional
|
||||
|
||||
import pytest
|
||||
import torch._dynamo
|
||||
@@ -41,8 +40,8 @@ FP8_DTYPE = current_platform.fp8_dtype()
|
||||
FP4_DTYPE = torch.uint8
|
||||
|
||||
# globals needed for string-import custom Dynamo backend field
|
||||
backend: Optional[TestBackend] = None
|
||||
backend_unfused: Optional[TestBackend] = None
|
||||
backend: TestBackend | None = None
|
||||
backend_unfused: TestBackend | None = None
|
||||
|
||||
|
||||
class AttentionQuantPatternModel(torch.nn.Module):
|
||||
|
||||
@@ -1,7 +1,6 @@
|
||||
# SPDX-License-Identifier: Apache-2.0
|
||||
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
|
||||
|
||||
from typing import Optional
|
||||
|
||||
import torch
|
||||
|
||||
@@ -10,7 +9,7 @@ from vllm.config import CompilationLevel
|
||||
|
||||
|
||||
class MyMod(torch.nn.Module):
|
||||
def forward(self, x: torch.Tensor, cache: Optional[torch.Tensor] = None):
|
||||
def forward(self, x: torch.Tensor, cache: torch.Tensor | None = None):
|
||||
if cache is not None:
|
||||
return x + cache
|
||||
return x * 2
|
||||
@@ -24,11 +23,11 @@ class MyWrapper(TorchCompileWrapperWithCustomDispatcher):
|
||||
compiled_callable, compilation_level=CompilationLevel.DYNAMO_ONCE
|
||||
)
|
||||
|
||||
def forward(self, x: torch.Tensor, cache: Optional[torch.Tensor] = None):
|
||||
def forward(self, x: torch.Tensor, cache: torch.Tensor | None = None):
|
||||
# this is the function to be compiled
|
||||
return self.model(x, cache)
|
||||
|
||||
def __call__(self, x: torch.Tensor, cache: Optional[torch.Tensor] = None):
|
||||
def __call__(self, x: torch.Tensor, cache: torch.Tensor | None = None):
|
||||
# let torch.compile compile twice
|
||||
if len(self.compiled_codes) == 2:
|
||||
dispatch_id = 0 if cache is None else 1
|
||||
|
||||
Reference in New Issue
Block a user