biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

Update Optional[x] -> x | None and Union[x, y] to x | y (#26633)

Browse Source 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Harry Mellor
2025-10-12 17:51:31 +01:00
committed by GitHub
parent 9bb38130cb
commit 8fcaaf6a16
944 changed files with 9490 additions and 10121 deletions
Download Patch File Download Diff File Expand all files Collapse all files

33
vllm/compilation/compiler_interface.py
View File

@@ -4,8 +4,9 @@ import contextlib
import copy
import hashlib
import os
from collections.abc import Callable
from contextlib import ExitStack
from typing import Any, Callable, Optional
from typing import Any
from unittest.mock import patch
import torch
@@ -62,9 +63,9 @@ class CompilerInterface:
graph: fx.GraphModule,
example_inputs: list[Any],
compiler_config: dict[str, Any],
runtime_shape: Optional[int] = None,
key: Optional[str] = None,
) -> tuple[Optional[Callable], Optional[Any]]:
runtime_shape: int | None = None,
key: str | None = None,
) -> tuple[Callable | None, Any | None]:
"""
Compile the graph with the given example inputs and compiler config,
with a runtime shape. If the `runtime_shape` is None, it means
@@ -97,7 +98,7 @@ class CompilerInterface:
graph: fx.GraphModule,
example_inputs: list[Any],
graph_index: int,
runtime_shape: Optional[int] = None,
runtime_shape: int | None = None,
) -> Callable:
"""
Load the compiled function from the handle.
@@ -191,9 +192,9 @@ class InductorStandaloneAdaptor(CompilerInterface):
graph: fx.GraphModule,
example_inputs: list[Any],
compiler_config: dict[str, Any],
runtime_shape: Optional[int] = None,
key: Optional[str] = None,
) -> tuple[Optional[Callable], Optional[Any]]:
runtime_shape: int | None = None,
key: str | None = None,
) -> tuple[Callable | None, Any | None]:
compilation_counter.num_inductor_compiles += 1
current_config = {}
if compiler_config is not None:
@@ -229,7 +230,7 @@ class InductorStandaloneAdaptor(CompilerInterface):
graph: fx.GraphModule,
example_inputs: list[Any],
graph_index: int,
runtime_shape: Optional[int] = None,
runtime_shape: int | None = None,
) -> Callable:
assert isinstance(handle, tuple)
assert isinstance(handle[0], str)
@@ -293,9 +294,9 @@ class InductorAdaptor(CompilerInterface):
graph: fx.GraphModule,
example_inputs: list[Any],
compiler_config: dict[str, Any],
runtime_shape: Optional[int] = None,
key: Optional[str] = None,
) -> tuple[Optional[Callable], Optional[Any]]:
runtime_shape: int | None = None,
key: str | None = None,
) -> tuple[Callable | None, Any | None]:
compilation_counter.num_inductor_compiles += 1
from torch._inductor.compile_fx import compile_fx
@@ -492,7 +493,7 @@ class InductorAdaptor(CompilerInterface):
graph: fx.GraphModule,
example_inputs: list[Any],
graph_index: int,
runtime_shape: Optional[int] = None,
runtime_shape: int | None = None,
) -> Callable:
assert isinstance(handle, tuple)
assert isinstance(handle[0], str)
@@ -610,9 +611,9 @@ class EagerAdaptor(CompilerInterface):
graph: fx.GraphModule,
example_inputs: list[Any],
compiler_config: dict[str, Any],
runtime_shape: Optional[int] = None,
key: Optional[str] = None,
) -> tuple[Optional[Callable], Optional[Any]]:
runtime_shape: int | None = None,
key: str | None = None,
) -> tuple[Callable | None, Any | None]:
compilation_counter.num_eager_compiles += 1
# we don't need to compile the graph, just return the graph itself.
# It does not support caching, return None for the handle.