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:
Harry Mellor
2025-10-12 09:51:31 -07:00
committed by GitHub
parent 9bb38130cb
commit 8fcaaf6a16
944 changed files with 9490 additions and 10121 deletions
@@ -11,10 +11,10 @@ import os
import tempfile
import time
from collections import defaultdict
from collections.abc import Generator
from collections.abc import Callable, Generator
from contextlib import contextmanager
from pathlib import Path
from typing import IO, Any, Callable, Optional, Union
from typing import IO, Any
import filelock
import huggingface_hub.constants
@@ -85,7 +85,7 @@ class DisabledTqdm(tqdm):
super().__init__(*args, **kwargs, disable=True)
def get_lock(model_name_or_path: Union[str, Path], cache_dir: Optional[str] = None):
def get_lock(model_name_or_path: str | Path, cache_dir: str | None = None):
lock_dir = cache_dir or temp_dir
model_name_or_path = str(model_name_or_path)
os.makedirs(os.path.dirname(lock_dir), exist_ok=True)
@@ -100,7 +100,7 @@ def get_lock(model_name_or_path: Union[str, Path], cache_dir: Optional[str] = No
@contextmanager
def atomic_writer(
filepath: Union[str, Path], mode: str = "w", encoding: Optional[str] = None
filepath: str | Path, mode: str = "w", encoding: str | None = None
) -> Generator[IO]:
"""
Context manager that provides an atomic file writing routine.
@@ -143,11 +143,11 @@ def atomic_writer(
def maybe_download_from_modelscope(
model: str,
revision: Optional[str] = None,
download_dir: Optional[str] = None,
ignore_patterns: Optional[Union[str, list[str]]] = None,
allow_patterns: Optional[Union[list[str], str]] = None,
) -> Optional[str]:
revision: str | None = None,
download_dir: str | None = None,
ignore_patterns: str | list[str] | None = None,
allow_patterns: list[str] | str | None = None,
) -> str | None:
"""Download model from ModelScope hub if VLLM_USE_MODELSCOPE is True.
Returns the path to the downloaded model, or None if the model is not
@@ -370,10 +370,10 @@ def get_sparse_attention_config(
def download_weights_from_hf(
model_name_or_path: str,
cache_dir: Optional[str],
cache_dir: str | None,
allow_patterns: list[str],
revision: Optional[str] = None,
ignore_patterns: Optional[Union[str, list[str]]] = None,
revision: str | None = None,
ignore_patterns: str | list[str] | None = None,
) -> str:
"""Download model weights from Hugging Face Hub.
@@ -448,8 +448,8 @@ def download_weights_from_hf(
def download_safetensors_index_file_from_hf(
model_name_or_path: str,
index_file: str,
cache_dir: Optional[str],
revision: Optional[str] = None,
cache_dir: str | None,
revision: str | None = None,
) -> None:
"""Download hf safetensors index file from Hugging Face Hub.
@@ -540,7 +540,7 @@ def enable_tqdm(use_tqdm_on_load: bool):
def np_cache_weights_iterator(
model_name_or_path: str,
cache_dir: Optional[str],
cache_dir: str | None,
hf_folder: str,
hf_weights_files: list[str],
use_tqdm_on_load: bool,
@@ -746,7 +746,7 @@ def fastsafetensors_weights_iterator(
def pt_weights_iterator(
hf_weights_files: list[str],
use_tqdm_on_load: bool,
pt_load_map_location: Union[str, dict[str, str]] = "cpu",
pt_load_map_location: str | dict[str, str] = "cpu",
) -> Generator[tuple[str, torch.Tensor], None, None]:
"""Iterate over the weights in the model bin/pt files."""
for bin_file in tqdm(
@@ -765,7 +765,7 @@ def pt_weights_iterator(
def multi_thread_pt_weights_iterator(
hf_weights_files: list[str],
use_tqdm_on_load: bool,
pt_load_map_location: Union[str, dict[str, str]] = "cpu",
pt_load_map_location: str | dict[str, str] = "cpu",
max_workers: int = 4,
) -> Generator[tuple[str, torch.Tensor], None, None]:
"""Multi-Thread iterate over the weights in the model bin/pt files."""
@@ -985,7 +985,7 @@ def initialize_dummy_weights(
param.uniform_(low, high, generator=generator)
def maybe_remap_kv_scale_name(name: str, params_dict: dict) -> Optional[str]:
def maybe_remap_kv_scale_name(name: str, params_dict: dict) -> str | None:
"""Remap the name of FP8 k/v_scale parameters.
This function handles the remapping of FP8 k/v_scale parameter names.