pytorch-ignite 0.6.0.dev20260115__tar.gz → 0.6.0.dev20260116__tar.gz

{pytorch_ignite-0.6.0.dev20260115 → pytorch_ignite-0.6.0.dev20260116}/.github/workflows/code-style-checks.yml

@@ -29,6 +29,7 @@ on:
       - "!docs/**"
       - "!conda.recipe"
   workflow_dispatch:
+  merge_group:
 
 concurrency:
   # <workflow_name>-<branch_name>-<true || commit_sha (if branch is protected)>

{pytorch_ignite-0.6.0.dev20260115 → pytorch_ignite-0.6.0.dev20260116}/.github/workflows/hvd-tests.yml

@@ -16,6 +16,7 @@ on:
       - "tests/run_cpu_tests.sh"
       - ".github/workflows/hvd-tests.yml"
   workflow_dispatch:
+  merge_group:
 
 concurrency:
   # <workflow_name>-<branch_name>-<true || commit_sha (if branch is protected)>

{pytorch_ignite-0.6.0.dev20260115 → pytorch_ignite-0.6.0.dev20260116}/.github/workflows/mps-tests.yml

@@ -18,6 +18,7 @@ on:
       - "requirements-dev.txt"
       - ".github/workflows/mps-tests.yml"
   workflow_dispatch:
+  merge_group:
 
 concurrency:
   # <workflow_name>-<branch_name>-<true || commit_sha (if branch is protected)>

{pytorch_ignite-0.6.0.dev20260115 → pytorch_ignite-0.6.0.dev20260116}/.github/workflows/typing-checks.yml

@@ -23,6 +23,7 @@ on:
       - "pyrefly.toml"
       - ".github/workflows/typing-checks.yml"
   workflow_dispatch:
+  merge_group:
 
 concurrency:
   # <workflow_name>-<branch_name>-<true || commit_sha (if branch is protected)>

{pytorch_ignite-0.6.0.dev20260115 → pytorch_ignite-0.6.0.dev20260116}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pytorch-ignite
-Version: 0.6.0.dev20260115
+Version: 0.6.0.dev20260116
 Summary: A lightweight library to help with training neural networks in PyTorch.
 Project-URL: Homepage, https://pytorch-ignite.ai
 Project-URL: Repository, https://github.com/pytorch/ignite

{pytorch_ignite-0.6.0.dev20260115 → pytorch_ignite-0.6.0.dev20260116}/conda.recipe/meta.yaml

@@ -1,4 +1,4 @@
-{% set __version__ = "0.6.0.dev20260115" %}
+{% set __version__ = "0.6.0.dev20260116" %}
 
 package:
   name: ignite

{pytorch_ignite-0.6.0.dev20260115 → pytorch_ignite-0.6.0.dev20260116}/ignite/__init__.py

@@ -6,4 +6,4 @@ import ignite.handlers
 import ignite.metrics
 import ignite.utils
 
-__version__ = "0.6.0.dev20260115"
+__version__ = "0.6.0.dev20260116"

{pytorch_ignite-0.6.0.dev20260115 → pytorch_ignite-0.6.0.dev20260116}/ignite/contrib/engines/common.py

@@ -1,7 +1,7 @@
 import numbers
 import warnings
 from functools import partial
-from typing import Any, Callable, cast, Dict, Iterable, Mapping, Optional, Sequence, Union
+from typing import Any, Callable, cast, Iterable, Mapping, Sequence
 
 import torch
 import torch.nn as nn
@@ -41,19 +41,19 @@ from ignite.utils import deprecated
 
 def setup_common_training_handlers(
     trainer: Engine,
-    train_sampler: Optional[DistributedSampler] = None,
-    to_save: Optional[Mapping] = None,
+    train_sampler: DistributedSampler | None = None,
+    to_save: Mapping | None = None,
     save_every_iters: int = 1000,
-    output_path: Optional[str] = None,
-    lr_scheduler: Optional[Union[ParamScheduler, PyTorchLRScheduler]] = None,
+    output_path: str | None = None,
+    lr_scheduler: ParamScheduler | PyTorchLRScheduler | None = None,
     with_gpu_stats: bool = False,
-    output_names: Optional[Iterable[str]] = None,
+    output_names: Iterable[str] | None = None,
     with_pbars: bool = True,
     with_pbar_on_iters: bool = True,
     log_every_iters: int = 100,
     stop_on_nan: bool = True,
     clear_cuda_cache: bool = True,
-    save_handler: Optional[Union[Callable, BaseSaveHandler]] = None,
+    save_handler: Callable | BaseSaveHandler | None = None,
     **kwargs: Any,
 ) -> None:
     """Helper method to setup trainer with common handlers (it also supports distributed configuration):
@@ -145,18 +145,18 @@ setup_common_distrib_training_handlers = setup_common_training_handlers
 
 def _setup_common_training_handlers(
     trainer: Engine,
-    to_save: Optional[Mapping] = None,
+    to_save: Mapping | None = None,
     save_every_iters: int = 1000,
-    output_path: Optional[str] = None,
-    lr_scheduler: Optional[Union[ParamScheduler, PyTorchLRScheduler]] = None,
+    output_path: str | None = None,
+    lr_scheduler: ParamScheduler | PyTorchLRScheduler | None = None,
     with_gpu_stats: bool = False,
-    output_names: Optional[Iterable[str]] = None,
+    output_names: Iterable[str] | None = None,
     with_pbars: bool = True,
     with_pbar_on_iters: bool = True,
     log_every_iters: int = 100,
     stop_on_nan: bool = True,
     clear_cuda_cache: bool = True,
-    save_handler: Optional[Union[Callable, BaseSaveHandler]] = None,
+    save_handler: Callable | BaseSaveHandler | None = None,
     **kwargs: Any,
 ) -> None:
     if output_path is not None and save_handler is not None:
@@ -185,7 +185,7 @@ def _setup_common_training_handlers(
             save_handler = DiskSaver(dirname=output_path, require_empty=False)
 
         checkpoint_handler = Checkpoint(
-            to_save, cast(Union[Callable, BaseSaveHandler], save_handler), filename_prefix="training", **kwargs
+            to_save, cast(Callable | BaseSaveHandler, save_handler), filename_prefix="training", **kwargs
         )
         trainer.add_event_handler(Events.ITERATION_COMPLETED(every=save_every_iters), checkpoint_handler)
 
@@ -227,19 +227,19 @@ def _setup_common_training_handlers(
 
 def _setup_common_distrib_training_handlers(
     trainer: Engine,
-    train_sampler: Optional[DistributedSampler] = None,
-    to_save: Optional[Mapping] = None,
+    train_sampler: DistributedSampler | None = None,
+    to_save: Mapping | None = None,
     save_every_iters: int = 1000,
-    output_path: Optional[str] = None,
-    lr_scheduler: Optional[Union[ParamScheduler, PyTorchLRScheduler]] = None,
+    output_path: str | None = None,
+    lr_scheduler: ParamScheduler | PyTorchLRScheduler | None = None,
     with_gpu_stats: bool = False,
-    output_names: Optional[Iterable[str]] = None,
+    output_names: Iterable[str] | None = None,
     with_pbars: bool = True,
     with_pbar_on_iters: bool = True,
     log_every_iters: int = 100,
     stop_on_nan: bool = True,
     clear_cuda_cache: bool = True,
-    save_handler: Optional[Union[Callable, BaseSaveHandler]] = None,
+    save_handler: Callable | BaseSaveHandler | None = None,
     **kwargs: Any,
 ) -> None:
     _setup_common_training_handlers(
@@ -286,8 +286,8 @@ def setup_any_logging(
     logger: BaseLogger,
     logger_module: Any,
     trainer: Engine,
-    optimizers: Optional[Union[Optimizer, Dict[str, Optimizer], Dict[None, Optimizer]]],
-    evaluators: Optional[Union[Engine, Dict[str, Engine]]],
+    optimizers: Optimizer | dict[str, Optimizer] | dict[None, Optimizer] | None,
+    evaluators: Engine | dict[str, Engine] | None,
     log_every_iters: int,
 ) -> None:
     pass
@@ -296,8 +296,8 @@ def setup_any_logging(
 def _setup_logging(
     logger: BaseLogger,
     trainer: Engine,
-    optimizers: Optional[Union[Optimizer, Dict[str, Optimizer], Dict[None, Optimizer]]],
-    evaluators: Optional[Union[Engine, Dict[str, Engine]]],
+    optimizers: Optimizer | dict[str, Optimizer] | dict[None, Optimizer] | None,
+    evaluators: Engine | dict[str, Engine] | None,
     log_every_iters: int,
 ) -> None:
     if optimizers is not None:
@@ -341,8 +341,8 @@ def _setup_logging(
 def setup_tb_logging(
     output_path: str,
     trainer: Engine,
-    optimizers: Optional[Union[Optimizer, Dict[str, Optimizer]]] = None,
-    evaluators: Optional[Union[Engine, Dict[str, Engine]]] = None,
+    optimizers: Optimizer | dict[str, Optimizer] | None = None,
+    evaluators: Engine | dict[str, Engine] | None = None,
     log_every_iters: int = 100,
     **kwargs: Any,
 ) -> TensorboardLogger:
@@ -373,8 +373,8 @@ def setup_tb_logging(
 
 def setup_visdom_logging(
     trainer: Engine,
-    optimizers: Optional[Union[Optimizer, Dict[str, Optimizer]]] = None,
-    evaluators: Optional[Union[Engine, Dict[str, Engine]]] = None,
+    optimizers: Optimizer | dict[str, Optimizer] | None = None,
+    evaluators: Engine | dict[str, Engine] | None = None,
     log_every_iters: int = 100,
     **kwargs: Any,
 ) -> VisdomLogger:
@@ -404,8 +404,8 @@ def setup_visdom_logging(
 
 def setup_mlflow_logging(
     trainer: Engine,
-    optimizers: Optional[Union[Optimizer, Dict[str, Optimizer]]] = None,
-    evaluators: Optional[Union[Engine, Dict[str, Engine]]] = None,
+    optimizers: Optimizer | dict[str, Optimizer] | None = None,
+    evaluators: Engine | dict[str, Engine] | None = None,
     log_every_iters: int = 100,
     **kwargs: Any,
 ) -> MLflowLogger:
@@ -435,8 +435,8 @@ def setup_mlflow_logging(
 
 def setup_neptune_logging(
     trainer: Engine,
-    optimizers: Optional[Union[Optimizer, Dict[str, Optimizer]]] = None,
-    evaluators: Optional[Union[Engine, Dict[str, Engine]]] = None,
+    optimizers: Optimizer | dict[str, Optimizer] | None = None,
+    evaluators: Engine | dict[str, Engine] | None = None,
     log_every_iters: int = 100,
     **kwargs: Any,
 ) -> NeptuneLogger:
@@ -466,8 +466,8 @@ def setup_neptune_logging(
 
 def setup_wandb_logging(
     trainer: Engine,
-    optimizers: Optional[Union[Optimizer, Dict[str, Optimizer]]] = None,
-    evaluators: Optional[Union[Engine, Dict[str, Engine]]] = None,
+    optimizers: Optimizer | dict[str, Optimizer] | None = None,
+    evaluators: Engine | dict[str, Engine] | None = None,
     log_every_iters: int = 100,
     **kwargs: Any,
 ) -> WandBLogger:
@@ -497,8 +497,8 @@ def setup_wandb_logging(
 
 def setup_plx_logging(
     trainer: Engine,
-    optimizers: Optional[Union[Optimizer, Dict[str, Optimizer]]] = None,
-    evaluators: Optional[Union[Engine, Dict[str, Engine]]] = None,
+    optimizers: Optimizer | dict[str, Optimizer] | None = None,
+    evaluators: Engine | dict[str, Engine] | None = None,
     log_every_iters: int = 100,
     **kwargs: Any,
 ) -> PolyaxonLogger:
@@ -528,8 +528,8 @@ def setup_plx_logging(
 
 def setup_clearml_logging(
     trainer: Engine,
-    optimizers: Optional[Union[Optimizer, Dict[str, Optimizer]]] = None,
-    evaluators: Optional[Union[Engine, Dict[str, Engine]]] = None,
+    optimizers: Optimizer | dict[str, Optimizer] | None = None,
+    evaluators: Engine | dict[str, Engine] | None = None,
     log_every_iters: int = 100,
     **kwargs: Any,
 ) -> ClearMLLogger:
@@ -559,8 +559,8 @@ def setup_clearml_logging(
 
 def setup_trains_logging(
     trainer: Engine,
-    optimizers: Optional[Union[Optimizer, Dict[str, Optimizer]]] = None,
-    evaluators: Optional[Union[Engine, Dict[str, Engine]]] = None,
+    optimizers: Optimizer | dict[str, Optimizer] | None = None,
+    evaluators: Engine | dict[str, Engine] | None = None,
     log_every_iters: int = 100,
     **kwargs: Any,
 ) -> ClearMLLogger:
@@ -573,12 +573,12 @@ get_default_score_fn = Checkpoint.get_default_score_fn
 
 
 def gen_save_best_models_by_val_score(
-    save_handler: Union[Callable, BaseSaveHandler],
+    save_handler: Callable | BaseSaveHandler,
     evaluator: Engine,
-    models: Union[torch.nn.Module, Dict[str, torch.nn.Module]],
+    models: torch.nn.Module | dict[str, torch.nn.Module],
     metric_name: str,
     n_saved: int = 3,
-    trainer: Optional[Engine] = None,
+    trainer: Engine | None = None,
     tag: str = "val",
     score_sign: float = 1.0,
     **kwargs: Any,
@@ -615,7 +615,7 @@ def gen_save_best_models_by_val_score(
         global_step_transform = global_step_from_engine(trainer)
 
     if isinstance(models, nn.Module):
-        to_save: Dict[str, nn.Module] = {"model": models}
+        to_save: dict[str, nn.Module] = {"model": models}
     else:
         to_save = models
 
@@ -640,7 +640,7 @@ def save_best_model_by_val_score(
     model: torch.nn.Module,
     metric_name: str,
     n_saved: int = 3,
-    trainer: Optional[Engine] = None,
+    trainer: Engine | None = None,
     tag: str = "val",
     score_sign: float = 1.0,
     **kwargs: Any,

{pytorch_ignite-0.6.0.dev20260115 → pytorch_ignite-0.6.0.dev20260116}/ignite/metrics/average_precision.py

@@ -1,4 +1,4 @@
-from typing import Callable, Union
+from typing import Callable
 
 import torch
 
@@ -71,7 +71,7 @@ class AveragePrecision(EpochMetric):
         self,
         output_transform: Callable = lambda x: x,
         check_compute_fn: bool = False,
-        device: Union[str, torch.device] = torch.device("cpu"),
+        device: str | torch.device = torch.device("cpu"),
         skip_unrolling: bool = False,
     ):
         try:

{pytorch_ignite-0.6.0.dev20260115 → pytorch_ignite-0.6.0.dev20260116}/ignite/metrics/cosine_similarity.py

@@ -1,4 +1,4 @@
-from typing import Callable, Sequence, Union
+from typing import Callable, Sequence
 
 import torch
 
@@ -77,7 +77,7 @@ class CosineSimilarity(Metric):
         self,
         eps: float = 1e-8,
         output_transform: Callable = lambda x: x,
-        device: Union[str, torch.device] = torch.device("cpu"),
+        device: str | torch.device = torch.device("cpu"),
         skip_unrolling: bool = False,
     ):
         super().__init__(output_transform, device, skip_unrolling=skip_unrolling)

{pytorch_ignite-0.6.0.dev20260115 → pytorch_ignite-0.6.0.dev20260116}/ignite/metrics/frequency.py

@@ -1,4 +1,4 @@
-from typing import Callable, Union
+from typing import Callable
 
 import torch
 
@@ -56,7 +56,7 @@ class Frequency(Metric):
     def __init__(
         self,
         output_transform: Callable = lambda x: x,
-        device: Union[str, torch.device] = torch.device("cpu"),
+        device: str | torch.device = torch.device("cpu"),
         skip_unrolling: bool = False,
     ) -> None:
         super(Frequency, self).__init__(output_transform=output_transform, device=device, skip_unrolling=skip_unrolling)

{pytorch_ignite-0.6.0.dev20260115 → pytorch_ignite-0.6.0.dev20260116}/ignite/metrics/mean_absolute_error.py

@@ -1,4 +1,4 @@
-from typing import Sequence, Union
+from typing import Sequence
 
 import torch
 
@@ -80,7 +80,7 @@ class MeanAbsoluteError(Metric):
         self._num_examples += y.shape[0]
 
     @sync_all_reduce("_sum_of_absolute_errors", "_num_examples")
-    def compute(self) -> Union[float, torch.Tensor]:
+    def compute(self) -> float | torch.Tensor:
         if self._num_examples == 0:
             raise NotComputableError("MeanAbsoluteError must have at least one example before it can be computed.")
         return self._sum_of_absolute_errors.item() / self._num_examples

{pytorch_ignite-0.6.0.dev20260115 → pytorch_ignite-0.6.0.dev20260116}/ignite/metrics/mean_squared_error.py

@@ -1,4 +1,4 @@
-from typing import Sequence, Union
+from typing import Sequence
 
 import torch
 
@@ -80,7 +80,7 @@ class MeanSquaredError(Metric):
         self._num_examples += y.shape[0]
 
     @sync_all_reduce("_sum_of_squared_errors", "_num_examples")
-    def compute(self) -> Union[float, torch.Tensor]:
+    def compute(self) -> float | torch.Tensor:
         if self._num_examples == 0:
             raise NotComputableError("MeanSquaredError must have at least one example before it can be computed.")
         return self._sum_of_squared_errors.item() / self._num_examples

{pytorch_ignite-0.6.0.dev20260115 → pytorch_ignite-0.6.0.dev20260116}/ignite/metrics/psnr.py

@@ -1,4 +1,4 @@
-from typing import Callable, Sequence, Union
+from typing import Callable, Sequence
 
 import torch
 
@@ -91,9 +91,9 @@ class PSNR(Metric):
 
     def __init__(
         self,
-        data_range: Union[int, float],
+        data_range: int | float,
         output_transform: Callable = lambda x: x,
-        device: Union[str, torch.device] = torch.device("cpu"),
+        device: str | torch.device = torch.device("cpu"),
         skip_unrolling: bool = False,
     ):
         super().__init__(output_transform=output_transform, device=device, skip_unrolling=skip_unrolling)

{pytorch_ignite-0.6.0.dev20260115 → pytorch_ignite-0.6.0.dev20260116}/ignite/metrics/root_mean_squared_error.py

@@ -1,5 +1,4 @@
 import math
-from typing import Union
 
 import torch
 
@@ -65,6 +64,6 @@ class RootMeanSquaredError(MeanSquaredError):
         ``skip_unrolling`` argument is added.
     """
 
-    def compute(self) -> Union[torch.Tensor, float]:
+    def compute(self) -> torch.Tensor | float:
         mse = super(RootMeanSquaredError, self).compute()
         return math.sqrt(mse)

{pytorch_ignite-0.6.0.dev20260115 → pytorch_ignite-0.6.0.dev20260116}/ignite/utils.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import collections.abc as collections
 import functools
 import hashlib
@@ -7,7 +9,7 @@ import random
 import shutil
 import warnings
 from pathlib import Path
-from typing import Any, Callable, cast, Dict, List, Optional, TextIO, Tuple, Type, TypeVar, Union
+from typing import Any, Callable, cast, TextIO, TypeVar
 
 import torch
 
@@ -24,10 +26,10 @@ __all__ = [
 
 
 def convert_tensor(
-    x: Union[torch.Tensor, collections.Sequence, collections.Mapping, str, bytes],
-    device: Optional[Union[str, torch.device]] = None,
+    x: torch.Tensor | collections.Sequence | collections.Mapping | str | bytes,
+    device: str | torch.device | None = None,
     non_blocking: bool = False,
-) -> Union[torch.Tensor, collections.Sequence, collections.Mapping, str, bytes]:
+) -> torch.Tensor | collections.Sequence | collections.Mapping | str | bytes:
     """Move tensors to relevant device.
 
     Args:
@@ -44,8 +46,8 @@ def convert_tensor(
 
 
 def apply_to_tensor(
-    x: Union[torch.Tensor, collections.Sequence, collections.Mapping, str, bytes], func: Callable
-) -> Union[torch.Tensor, collections.Sequence, collections.Mapping, str, bytes]:
+    x: torch.Tensor | collections.Sequence | collections.Mapping | str | bytes, func: Callable
+) -> torch.Tensor | collections.Sequence | collections.Mapping | str | bytes:
     """Apply a function on a tensor or mapping, or sequence of tensors.
 
     Args:
@@ -56,10 +58,10 @@ def apply_to_tensor(
 
 
 def apply_to_type(
-    x: Union[Any, collections.Sequence, collections.Mapping, str, bytes],
-    input_type: Union[Type, Tuple[Type[Any], Any]],
+    x: Any | collections.Sequence | collections.Mapping | str | bytes,
+    input_type: type | tuple[type[Any], Any],
     func: Callable,
-) -> Union[Any, collections.Sequence, collections.Mapping, str, bytes]:
+) -> Any | collections.Sequence | collections.Mapping | str | bytes:
     """Apply a function on an object of `input_type` or mapping, or sequence of objects of `input_type`.
 
     Args:
@@ -81,8 +83,8 @@ def apply_to_type(
 
 
 def _tree_map(
-    func: Callable, x: Union[Any, collections.Sequence, collections.Mapping], key: Optional[Union[int, str]] = None
-) -> Union[Any, collections.Sequence, collections.Mapping]:
+    func: Callable, x: Any | collections.Sequence | collections.Mapping, key: int | str | None = None
+) -> Any | collections.Sequence | collections.Mapping:
     if isinstance(x, collections.Mapping):
         return cast(Callable, type(x))({k: _tree_map(func, sample, key=k) for k, sample in x.items()})
     if isinstance(x, tuple) and hasattr(x, "_fields"):  # namedtuple
@@ -92,7 +94,7 @@ def _tree_map(
     return func(x, key=key)
 
 
-def _to_str_list(data: Any) -> List[str]:
+def _to_str_list(data: Any) -> list[str]:
     """
     Recursively flattens and formats complex data structures, including keys for
     dictionaries, into a list of human-readable strings.
@@ -127,9 +129,9 @@ def _to_str_list(data: Any) -> List[str]:
         A list of formatted strings, each representing a part of the input data
         structure.
     """
-    formatted_items: List[str] = []
+    formatted_items: list[str] = []
 
-    def format_item(item: Any, prefix: str = "") -> Optional[str]:
+    def format_item(item: Any, prefix: str = "") -> str | None:
         if isinstance(item, numbers.Number):
             return f"{prefix}{item:.4f}"
         elif torch.is_tensor(item):
@@ -169,9 +171,9 @@ def _to_str_list(data: Any) -> List[str]:
 
 
 class _CollectionItem:
-    types_as_collection_item: Tuple = (int, float, torch.Tensor)
+    types_as_collection_item: tuple = (int, float, torch.Tensor)
 
-    def __init__(self, collection: Union[Dict, List], key: Union[int, str]) -> None:
+    def __init__(self, collection: dict | list, key: int | str) -> None:
         if not isinstance(collection, (dict, list)):
             raise TypeError(
                 f"Input type is expected to be a mapping or list, but got {type(collection)} " f"for input key '{key}'."
@@ -189,7 +191,7 @@ class _CollectionItem:
         return self.collection[self.key]  # type: ignore[index]
 
     @staticmethod
-    def wrap(object: Union[Dict, List], key: Union[int, str], value: Any) -> Union[Any, "_CollectionItem"]:
+    def wrap(object: dict | list, key: int | str, value: Any) -> Any | "_CollectionItem":
         return (
             _CollectionItem(object, key)
             if value is None or isinstance(value, _CollectionItem.types_as_collection_item)
@@ -199,8 +201,8 @@ class _CollectionItem:
 
 def _tree_apply2(
     func: Callable,
-    x: Union[Any, List, Dict],
-    y: Union[Any, collections.Sequence, collections.Mapping],
+    x: Any | list | dict,
+    y: Any | collections.Sequence | collections.Mapping,
 ) -> None:
     if isinstance(x, dict) and isinstance(y, collections.Mapping):
         for k, v in x.items():
@@ -234,14 +236,14 @@ def to_onehot(indices: torch.Tensor, num_classes: int) -> torch.Tensor:
 
 
 def setup_logger(
-    name: Optional[str] = "ignite",
+    name: str | None = "ignite",
     level: int = logging.INFO,
-    stream: Optional[TextIO] = None,
+    stream: TextIO | None = None,
     format: str = "%(asctime)s %(name)s %(levelname)s: %(message)s",
-    filepath: Optional[str] = None,
-    distributed_rank: Optional[int] = None,
+    filepath: str | None = None,
+    distributed_rank: int | None = None,
     reset: bool = False,
-    encoding: Optional[str] = "utf-8",
+    encoding: str | None = "utf-8",
 ) -> logging.Logger:
     """Setups logger: name, level, format etc.
 
@@ -393,7 +395,7 @@ def manual_seed(seed: int) -> None:
 
 
 def deprecated(
-    deprecated_in: str, removed_in: str = "", reasons: Tuple[str, ...] = (), raise_exception: bool = False
+    deprecated_in: str, removed_in: str = "", reasons: tuple[str, ...] = (), raise_exception: bool = False
 ) -> Callable:
     F = TypeVar("F", bound=Callable[..., Any])
 
@@ -406,7 +408,7 @@ def deprecated(
         )
 
         @functools.wraps(func)
-        def wrapper(*args: Any, **kwargs: Dict[str, Any]) -> Callable:
+        def wrapper(*args: Any, **kwargs: dict[str, Any]) -> Callable:
             if raise_exception:
                 raise DeprecationWarning(deprecation_warning)
             warnings.warn(deprecation_warning, DeprecationWarning, stacklevel=2)
@@ -422,7 +424,7 @@ def deprecated(
     return decorator
 
 
-def hash_checkpoint(checkpoint_path: Union[str, Path], output_dir: Union[str, Path]) -> Tuple[Path, str]:
+def hash_checkpoint(checkpoint_path: str | Path, output_dir: str | Path) -> tuple[Path, str]:
     """
     Hash the checkpoint file in the format of ``<filename>-<hash>.<ext>``
     to be used with ``check_hash`` of :func:`torch.hub.load_state_dict_from_url`.