dgenerate-ultralytics-headless 8.3.190__py3-none-any.whl → 8.3.191__py3-none-any.whl

ultralytics/models/yolo/obb/val.py

@@ -1,7 +1,9 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
+from __future__ import annotations
+
 from pathlib import Path
-from typing import Any, Dict, List, Tuple
+from typing import Any
 
 import numpy as np
 import torch
@@ -70,7 +72,7 @@ class OBBValidator(DetectionValidator):
         self.is_dota = isinstance(val, str) and "DOTA" in val  # check if dataset is DOTA format
         self.confusion_matrix.task = "obb"  # set confusion matrix task to 'obb'
 
-    def _process_batch(self, preds: Dict[str, torch.Tensor], batch: Dict[str, torch.Tensor]) -> Dict[str, np.ndarray]:
+    def _process_batch(self, preds: dict[str, torch.Tensor], batch: dict[str, torch.Tensor]) -> dict[str, np.ndarray]:
         """
         Compute the correct prediction matrix for a batch of detections and ground truth bounding boxes.
 
@@ -96,7 +98,7 @@ class OBBValidator(DetectionValidator):
         iou = batch_probiou(batch["bboxes"], preds["bboxes"])
         return {"tp": self.match_predictions(preds["cls"], batch["cls"], iou).cpu().numpy()}
 
-    def postprocess(self, preds: torch.Tensor) -> List[Dict[str, torch.Tensor]]:
+    def postprocess(self, preds: torch.Tensor) -> list[dict[str, torch.Tensor]]:
         """
         Args:
             preds (torch.Tensor): Raw predictions from the model.
@@ -109,7 +111,7 @@ class OBBValidator(DetectionValidator):
             pred["bboxes"] = torch.cat([pred["bboxes"], pred.pop("extra")], dim=-1)  # concatenate angle
         return preds
 
-    def _prepare_batch(self, si: int, batch: Dict[str, Any]) -> Dict[str, Any]:
+    def _prepare_batch(self, si: int, batch: dict[str, Any]) -> dict[str, Any]:
         """
         Prepare batch data for OBB validation with proper scaling and formatting.
 
@@ -143,7 +145,7 @@ class OBBValidator(DetectionValidator):
             "im_file": batch["im_file"][si],
         }
 
-    def plot_predictions(self, batch: Dict[str, Any], preds: List[torch.Tensor], ni: int) -> None:
+    def plot_predictions(self, batch: dict[str, Any], preds: list[torch.Tensor], ni: int) -> None:
         """
         Plot predicted bounding boxes on input images and save the result.
 
@@ -163,7 +165,7 @@ class OBBValidator(DetectionValidator):
             p["bboxes"][:, :4] = ops.xywh2xyxy(p["bboxes"][:, :4])  # convert to xyxy format for plotting
         super().plot_predictions(batch, preds, ni)  # plot bboxes
 
-    def pred_to_json(self, predn: Dict[str, torch.Tensor], pbatch: Dict[str, Any]) -> None:
+    def pred_to_json(self, predn: dict[str, torch.Tensor], pbatch: dict[str, Any]) -> None:
         """
         Convert YOLO predictions to COCO JSON format with rotated bounding box information.
 
@@ -194,7 +196,7 @@ class OBBValidator(DetectionValidator):
                 }
             )
 
-    def save_one_txt(self, predn: Dict[str, torch.Tensor], save_conf: bool, shape: Tuple[int, int], file: Path) -> None:
+    def save_one_txt(self, predn: dict[str, torch.Tensor], save_conf: bool, shape: tuple[int, int], file: Path) -> None:
         """
         Save YOLO OBB detections to a text file in normalized coordinates.
 
@@ -221,7 +223,7 @@ class OBBValidator(DetectionValidator):
             obb=torch.cat([predn["bboxes"], predn["conf"].unsqueeze(-1), predn["cls"].unsqueeze(-1)], dim=1),
         ).save_txt(file, save_conf=save_conf)
 
-    def scale_preds(self, predn: Dict[str, torch.Tensor], pbatch: Dict[str, Any]) -> Dict[str, torch.Tensor]:
+    def scale_preds(self, predn: dict[str, torch.Tensor], pbatch: dict[str, Any]) -> dict[str, torch.Tensor]:
         """Scales predictions to the original image size."""
         return {
             **predn,
@@ -230,7 +232,7 @@ class OBBValidator(DetectionValidator):
             ),
         }
 
-    def eval_json(self, stats: Dict[str, Any]) -> Dict[str, Any]:
+    def eval_json(self, stats: dict[str, Any]) -> dict[str, Any]:
         """
         Evaluate YOLO output in JSON format and save predictions in DOTA format.
 

ultralytics/models/yolo/pose/train.py

@@ -1,8 +1,10 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
+from __future__ import annotations
+
 from copy import copy
 from pathlib import Path
-from typing import Any, Dict, Optional, Union
+from typing import Any
 
 from ultralytics.models import yolo
 from ultralytics.nn.tasks import PoseModel
@@ -38,7 +40,7 @@ class PoseTrainer(yolo.detect.DetectionTrainer):
         >>> trainer.train()
     """
 
-    def __init__(self, cfg=DEFAULT_CFG, overrides: Optional[Dict[str, Any]] = None, _callbacks=None):
+    def __init__(self, cfg=DEFAULT_CFG, overrides: dict[str, Any] | None = None, _callbacks=None):
         """
         Initialize a PoseTrainer object for training YOLO pose estimation models.
 
@@ -73,8 +75,8 @@ class PoseTrainer(yolo.detect.DetectionTrainer):
 
     def get_model(
         self,
-        cfg: Optional[Union[str, Path, Dict[str, Any]]] = None,
-        weights: Optional[Union[str, Path]] = None,
+        cfg: str | Path | dict[str, Any] | None = None,
+        weights: str | Path | None = None,
         verbose: bool = True,
     ) -> PoseModel:
         """
@@ -112,7 +114,7 @@ class PoseTrainer(yolo.detect.DetectionTrainer):
         """Plot training/validation metrics."""
         plot_results(file=self.csv, pose=True, on_plot=self.on_plot)  # save results.png
 
-    def get_dataset(self) -> Dict[str, Any]:
+    def get_dataset(self) -> dict[str, Any]:
         """
         Retrieve the dataset and ensure it contains the required `kpt_shape` key.
 

ultralytics/models/yolo/pose/val.py

@@ -1,7 +1,9 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
+from __future__ import annotations
+
 from pathlib import Path
-from typing import Any, Dict, Tuple
+from typing import Any
 
 import numpy as np
 import torch
@@ -81,7 +83,7 @@ class PoseValidator(DetectionValidator):
                 "See https://github.com/ultralytics/ultralytics/issues/4031."
             )
 
-    def preprocess(self, batch: Dict[str, Any]) -> Dict[str, Any]:
+    def preprocess(self, batch: dict[str, Any]) -> dict[str, Any]:
         """Preprocess batch by converting keypoints data to float and moving it to the device."""
         batch = super().preprocess(batch)
         batch["keypoints"] = batch["keypoints"].to(self.device).float()
@@ -116,7 +118,7 @@ class PoseValidator(DetectionValidator):
         nkpt = self.kpt_shape[0]
         self.sigma = OKS_SIGMA if is_pose else np.ones(nkpt) / nkpt
 
-    def postprocess(self, preds: torch.Tensor) -> Dict[str, torch.Tensor]:
+    def postprocess(self, preds: torch.Tensor) -> dict[str, torch.Tensor]:
         """
         Postprocess YOLO predictions to extract and reshape keypoints for pose estimation.
 
@@ -146,7 +148,7 @@ class PoseValidator(DetectionValidator):
             pred["keypoints"] = pred.pop("extra").view(-1, *self.kpt_shape)  # remove extra if exists
         return preds
 
-    def _prepare_batch(self, si: int, batch: Dict[str, Any]) -> Dict[str, Any]:
+    def _prepare_batch(self, si: int, batch: dict[str, Any]) -> dict[str, Any]:
         """
         Prepare a batch for processing by converting keypoints to float and scaling to original dimensions.
 
@@ -170,7 +172,7 @@ class PoseValidator(DetectionValidator):
         pbatch["keypoints"] = kpts
         return pbatch
 
-    def _process_batch(self, preds: Dict[str, torch.Tensor], batch: Dict[str, Any]) -> Dict[str, np.ndarray]:
+    def _process_batch(self, preds: dict[str, torch.Tensor], batch: dict[str, Any]) -> dict[str, np.ndarray]:
         """
         Return correct prediction matrix by computing Intersection over Union (IoU) between detections and ground truth.
 
@@ -200,7 +202,7 @@ class PoseValidator(DetectionValidator):
         tp.update({"tp_p": tp_p})  # update tp with kpts IoU
         return tp
 
-    def save_one_txt(self, predn: Dict[str, torch.Tensor], save_conf: bool, shape: Tuple[int, int], file: Path) -> None:
+    def save_one_txt(self, predn: dict[str, torch.Tensor], save_conf: bool, shape: tuple[int, int], file: Path) -> None:
         """
         Save YOLO pose detections to a text file in normalized coordinates.
 
@@ -224,7 +226,7 @@ class PoseValidator(DetectionValidator):
             keypoints=predn["keypoints"],
         ).save_txt(file, save_conf=save_conf)
 
-    def pred_to_json(self, predn: Dict[str, torch.Tensor], pbatch: Dict[str, Any]) -> None:
+    def pred_to_json(self, predn: dict[str, torch.Tensor], pbatch: dict[str, Any]) -> None:
         """
         Convert YOLO predictions to COCO JSON format.
 
@@ -246,7 +248,7 @@ class PoseValidator(DetectionValidator):
         for i, k in enumerate(kpts.flatten(1, 2).tolist()):
             self.jdict[-len(kpts) + i]["keypoints"] = k  # keypoints
 
-    def scale_preds(self, predn: Dict[str, torch.Tensor], pbatch: Dict[str, Any]) -> Dict[str, torch.Tensor]:
+    def scale_preds(self, predn: dict[str, torch.Tensor], pbatch: dict[str, Any]) -> dict[str, torch.Tensor]:
         """Scales predictions to the original image size."""
         return {
             **super().scale_preds(predn, pbatch),
@@ -258,7 +260,7 @@ class PoseValidator(DetectionValidator):
             ),
         }
 
-    def eval_json(self, stats: Dict[str, Any]) -> Dict[str, Any]:
+    def eval_json(self, stats: dict[str, Any]) -> dict[str, Any]:
         """Evaluate object detection model using COCO JSON format."""
         anno_json = self.data["path"] / "annotations/person_keypoints_val2017.json"  # annotations
         pred_json = self.save_dir / "predictions.json"  # predictions

ultralytics/models/yolo/segment/train.py

@@ -1,8 +1,9 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
+from __future__ import annotations
+
 from copy import copy
 from pathlib import Path
-from typing import Dict, Optional, Union
 
 from ultralytics.models import yolo
 from ultralytics.nn.tasks import SegmentationModel
@@ -27,7 +28,7 @@ class SegmentationTrainer(yolo.detect.DetectionTrainer):
         >>> trainer.train()
     """
 
-    def __init__(self, cfg=DEFAULT_CFG, overrides: Optional[Dict] = None, _callbacks=None):
+    def __init__(self, cfg=DEFAULT_CFG, overrides: dict | None = None, _callbacks=None):
         """
         Initialize a SegmentationTrainer object.
 
@@ -50,9 +51,7 @@ class SegmentationTrainer(yolo.detect.DetectionTrainer):
         overrides["task"] = "segment"
         super().__init__(cfg, overrides, _callbacks)
 
-    def get_model(
-        self, cfg: Optional[Union[Dict, str]] = None, weights: Optional[Union[str, Path]] = None, verbose: bool = True
-    ):
+    def get_model(self, cfg: dict | str | None = None, weights: str | Path | None = None, verbose: bool = True):
         """
         Initialize and return a SegmentationModel with specified configuration and weights.
 

ultralytics/models/yolo/segment/val.py

@@ -1,8 +1,10 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
+from __future__ import annotations
+
 from multiprocessing.pool import ThreadPool
 from pathlib import Path
-from typing import Any, Dict, List, Tuple
+from typing import Any
 
 import numpy as np
 import torch
@@ -50,7 +52,7 @@ class SegmentationValidator(DetectionValidator):
         self.args.task = "segment"
         self.metrics = SegmentMetrics()
 
-    def preprocess(self, batch: Dict[str, Any]) -> Dict[str, Any]:
+    def preprocess(self, batch: dict[str, Any]) -> dict[str, Any]:
         """
         Preprocess batch of images for YOLO segmentation validation.
 
@@ -93,7 +95,7 @@ class SegmentationValidator(DetectionValidator):
             "mAP50-95)",
         )
 
-    def postprocess(self, preds: List[torch.Tensor]) -> List[Dict[str, torch.Tensor]]:
+    def postprocess(self, preds: list[torch.Tensor]) -> list[dict[str, torch.Tensor]]:
         """
         Post-process YOLO predictions and return output detections with proto.
 
@@ -119,7 +121,7 @@ class SegmentationValidator(DetectionValidator):
             )
         return preds
 
-    def _prepare_batch(self, si: int, batch: Dict[str, Any]) -> Dict[str, Any]:
+    def _prepare_batch(self, si: int, batch: dict[str, Any]) -> dict[str, Any]:
         """
         Prepare a batch for training or inference by processing images and targets.
 
@@ -135,7 +137,7 @@ class SegmentationValidator(DetectionValidator):
         prepared_batch["masks"] = batch["masks"][midx]
         return prepared_batch
 
-    def _process_batch(self, preds: Dict[str, torch.Tensor], batch: Dict[str, Any]) -> Dict[str, np.ndarray]:
+    def _process_batch(self, preds: dict[str, torch.Tensor], batch: dict[str, Any]) -> dict[str, np.ndarray]:
         """
         Compute correct prediction matrix for a batch based on bounding boxes and optional masks.
 
@@ -174,7 +176,7 @@ class SegmentationValidator(DetectionValidator):
         tp.update({"tp_m": tp_m})  # update tp with mask IoU
         return tp
 
-    def plot_predictions(self, batch: Dict[str, Any], preds: List[Dict[str, torch.Tensor]], ni: int) -> None:
+    def plot_predictions(self, batch: dict[str, Any], preds: list[dict[str, torch.Tensor]], ni: int) -> None:
         """
         Plot batch predictions with masks and bounding boxes.
 
@@ -190,7 +192,7 @@ class SegmentationValidator(DetectionValidator):
             p["masks"] = torch.as_tensor(masks[:50], dtype=torch.uint8).cpu()
         super().plot_predictions(batch, preds, ni, max_det=50)  # plot bboxes
 
-    def save_one_txt(self, predn: torch.Tensor, save_conf: bool, shape: Tuple[int, int], file: Path) -> None:
+    def save_one_txt(self, predn: torch.Tensor, save_conf: bool, shape: tuple[int, int], file: Path) -> None:
         """
         Save YOLO detections to a txt file in normalized coordinates in a specific format.
 
@@ -210,7 +212,7 @@ class SegmentationValidator(DetectionValidator):
             masks=torch.as_tensor(predn["masks"], dtype=torch.uint8),
         ).save_txt(file, save_conf=save_conf)
 
-    def pred_to_json(self, predn: Dict[str, torch.Tensor], pbatch: Dict[str, Any]) -> None:
+    def pred_to_json(self, predn: dict[str, torch.Tensor], pbatch: dict[str, Any]) -> None:
         """
         Save one JSON result for COCO evaluation.
 
@@ -233,7 +235,7 @@ class SegmentationValidator(DetectionValidator):
         for i, r in enumerate(rles):
             self.jdict[-len(rles) + i]["segmentation"] = r  # segmentation
 
-    def scale_preds(self, predn: Dict[str, torch.Tensor], pbatch: Dict[str, Any]) -> Dict[str, torch.Tensor]:
+    def scale_preds(self, predn: dict[str, torch.Tensor], pbatch: dict[str, Any]) -> dict[str, torch.Tensor]:
         """Scales predictions to the original image size."""
         return {
             **super().scale_preds(predn, pbatch),
@@ -244,7 +246,7 @@ class SegmentationValidator(DetectionValidator):
             ),
         }
 
-    def eval_json(self, stats: Dict[str, Any]) -> Dict[str, Any]:
+    def eval_json(self, stats: dict[str, Any]) -> dict[str, Any]:
         """Return COCO-style instance segmentation evaluation metrics."""
         pred_json = self.save_dir / "predictions.json"  # predictions
         anno_json = (

ultralytics/models/yolo/world/train.py

@@ -1,8 +1,10 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
+from __future__ import annotations
+
 import itertools
 from pathlib import Path
-from typing import Any, Dict, List, Optional
+from typing import Any
 
 import torch
 
@@ -51,7 +53,7 @@ class WorldTrainer(DetectionTrainer):
         >>> trainer.train()
     """
 
-    def __init__(self, cfg=DEFAULT_CFG, overrides: Optional[Dict[str, Any]] = None, _callbacks=None):
+    def __init__(self, cfg=DEFAULT_CFG, overrides: dict[str, Any] | None = None, _callbacks=None):
         """
         Initialize a WorldTrainer object with given arguments.
 
@@ -65,7 +67,7 @@ class WorldTrainer(DetectionTrainer):
         super().__init__(cfg, overrides, _callbacks)
         self.text_embeddings = None
 
-    def get_model(self, cfg=None, weights: Optional[str] = None, verbose: bool = True) -> WorldModel:
+    def get_model(self, cfg=None, weights: str | None = None, verbose: bool = True) -> WorldModel:
         """
         Return WorldModel initialized with specified config and weights.
 
@@ -91,7 +93,7 @@ class WorldTrainer(DetectionTrainer):
 
         return model
 
-    def build_dataset(self, img_path: str, mode: str = "train", batch: Optional[int] = None):
+    def build_dataset(self, img_path: str, mode: str = "train", batch: int | None = None):
         """
         Build YOLO Dataset for training or validation.
 
@@ -111,7 +113,7 @@ class WorldTrainer(DetectionTrainer):
             self.set_text_embeddings([dataset], batch)  # cache text embeddings to accelerate training
         return dataset
 
-    def set_text_embeddings(self, datasets: List[Any], batch: Optional[int]) -> None:
+    def set_text_embeddings(self, datasets: list[Any], batch: int | None) -> None:
         """
         Set text embeddings for datasets to accelerate training by caching category names.
 
@@ -137,7 +139,7 @@ class WorldTrainer(DetectionTrainer):
             )
         self.text_embeddings = text_embeddings
 
-    def generate_text_embeddings(self, texts: List[str], batch: int, cache_dir: Path) -> Dict[str, torch.Tensor]:
+    def generate_text_embeddings(self, texts: list[str], batch: int, cache_dir: Path) -> dict[str, torch.Tensor]:
         """
         Generate text embeddings for a list of text samples.
 
@@ -163,7 +165,7 @@ class WorldTrainer(DetectionTrainer):
         torch.save(txt_map, cache_path)
         return txt_map
 
-    def preprocess_batch(self, batch: Dict[str, Any]) -> Dict[str, Any]:
+    def preprocess_batch(self, batch: dict[str, Any]) -> dict[str, Any]:
         """Preprocess a batch of images and text for YOLOWorld training."""
         batch = DetectionTrainer.preprocess_batch(self, batch)
 

ultralytics/models/yolo/yoloe/train.py

@@ -1,9 +1,10 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
+from __future__ import annotations
+
 import itertools
 from copy import copy, deepcopy
 from pathlib import Path
-from typing import Dict, List, Optional, Union
 
 import torch
 
@@ -34,7 +35,7 @@ class YOLOETrainer(DetectionTrainer):
         build_dataset: Build YOLO dataset with multi-modal support for training.
     """
 
-    def __init__(self, cfg=DEFAULT_CFG, overrides: Optional[Dict] = None, _callbacks=None):
+    def __init__(self, cfg=DEFAULT_CFG, overrides: dict | None = None, _callbacks=None):
         """
         Initialize the YOLOE Trainer with specified configurations.
 
@@ -89,7 +90,7 @@ class YOLOETrainer(DetectionTrainer):
             self.test_loader, save_dir=self.save_dir, args=copy(self.args), _callbacks=self.callbacks
         )
 
-    def build_dataset(self, img_path: str, mode: str = "train", batch: Optional[int] = None):
+    def build_dataset(self, img_path: str, mode: str = "train", batch: int | None = None):
         """
         Build YOLO Dataset.
 
@@ -174,7 +175,7 @@ class YOLOETrainerFromScratch(YOLOETrainer, WorldTrainerFromScratch):
         generate_text_embeddings: Generate and cache text embeddings for training.
     """
 
-    def build_dataset(self, img_path: Union[List[str], str], mode: str = "train", batch: Optional[int] = None):
+    def build_dataset(self, img_path: list[str] | str, mode: str = "train", batch: int | None = None):
         """
         Build YOLO Dataset for training or validation.
 
@@ -201,7 +202,7 @@ class YOLOETrainerFromScratch(YOLOETrainer, WorldTrainerFromScratch):
         batch["txt_feats"] = txt_feats
         return batch
 
-    def generate_text_embeddings(self, texts: List[str], batch: int, cache_dir: Path):
+    def generate_text_embeddings(self, texts: list[str], batch: int, cache_dir: Path):
         """
         Generate text embeddings for a list of text samples.
 
@@ -285,7 +286,7 @@ class YOLOEVPTrainer(YOLOETrainerFromScratch):
         preprocess_batch: Preprocess batches with visual prompts.
     """
 
-    def build_dataset(self, img_path: Union[List[str], str], mode: str = "train", batch: Optional[int] = None):
+    def build_dataset(self, img_path: list[str] | str, mode: str = "train", batch: int | None = None):
         """
         Build YOLO Dataset for training or validation with visual prompts.
 

ultralytics/models/yolo/yoloe/val.py

@@ -1,8 +1,10 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
+from __future__ import annotations
+
 from copy import deepcopy
 from pathlib import Path
-from typing import Any, Dict, Optional, Union
+from typing import Any
 
 import torch
 from torch.nn import functional as F
@@ -96,14 +98,14 @@ class YOLOEDetectValidator(DetectionValidator):
         visual_pe[cls_visual_num == 0] = 0
         return visual_pe.unsqueeze(0)
 
-    def preprocess(self, batch: Dict[str, Any]) -> Dict[str, Any]:
+    def preprocess(self, batch: dict[str, Any]) -> dict[str, Any]:
         """Preprocess batch data, ensuring visuals are on the same device as images."""
         batch = super().preprocess(batch)
         if "visuals" in batch:
             batch["visuals"] = batch["visuals"].to(batch["img"].device)
         return batch
 
-    def get_vpe_dataloader(self, data: Dict[str, Any]) -> torch.utils.data.DataLoader:
+    def get_vpe_dataloader(self, data: dict[str, Any]) -> torch.utils.data.DataLoader:
         """
         Create a dataloader for LVIS training visual prompt samples.
 
@@ -141,11 +143,11 @@ class YOLOEDetectValidator(DetectionValidator):
     @smart_inference_mode()
     def __call__(
         self,
-        trainer: Optional[Any] = None,
-        model: Optional[Union[YOLOEModel, str]] = None,
-        refer_data: Optional[str] = None,
+        trainer: Any | None = None,
+        model: YOLOEModel | str | None = None,
+        refer_data: str | None = None,
         load_vp: bool = False,
-    ) -> Dict[str, Any]:
+    ) -> dict[str, Any]:
         """
         Run validation on the model using either text or visual prompt embeddings.
 
@@ -186,7 +188,7 @@ class YOLOEDetectValidator(DetectionValidator):
             if isinstance(model, (str, Path)):
                 from ultralytics.nn.tasks import attempt_load_weights
 
-                model = attempt_load_weights(model, device=self.device, inplace=True)
+                model = attempt_load_weights(model, device=self.device)
             model.eval().to(self.device)
             data = check_det_dataset(refer_data or self.args.data)
             names = [name.split("/", 1)[0] for name in list(data["names"].values())]

ultralytics/nn/autobackend.py

@@ -1,12 +1,14 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
+from __future__ import annotations
+
 import ast
 import json
 import platform
 import zipfile
 from collections import OrderedDict, namedtuple
 from pathlib import Path
-from typing import Any, Dict, List, Optional, Tuple, Union
+from typing import Any
 
 import cv2
 import numpy as np
@@ -19,7 +21,7 @@ from ultralytics.utils.checks import check_requirements, check_suffix, check_ver
 from ultralytics.utils.downloads import attempt_download_asset, is_url
 
 
-def check_class_names(names: Union[List, Dict]) -> Dict[int, str]:
+def check_class_names(names: list | dict) -> dict[int, str]:
     """
     Check class names and convert to dict format if needed.
 
@@ -49,7 +51,7 @@ def check_class_names(names: Union[List, Dict]) -> Dict[int, str]:
     return names
 
 
-def default_class_names(data: Optional[Union[str, Path]] = None) -> Dict[int, str]:
+def default_class_names(data: str | Path | None = None) -> dict[int, str]:
     """
     Apply default class names to an input YAML file or return numerical class names.
 
@@ -134,10 +136,10 @@ class AutoBackend(nn.Module):
     @torch.no_grad()
     def __init__(
         self,
-        model: Union[str, List[str], torch.nn.Module] = "yolo11n.pt",
+        model: str | torch.nn.Module = "yolo11n.pt",
         device: torch.device = torch.device("cpu"),
         dnn: bool = False,
-        data: Optional[Union[str, Path]] = None,
+        data: str | Path | None = None,
         fp16: bool = False,
         fuse: bool = True,
         verbose: bool = True,
@@ -146,7 +148,7 @@ class AutoBackend(nn.Module):
         Initialize the AutoBackend for inference.
 
         Args:
-            model (str | List[str] | torch.nn.Module): Path to the model weights file or a module instance.
+            model (str | torch.nn.Module): Path to the model weights file or a module instance.
             device (torch.device): Device to run the model on.
             dnn (bool): Use OpenCV DNN module for ONNX inference.
             data (str | Path, optional): Path to the additional data.yaml file containing class names.
@@ -155,7 +157,6 @@ class AutoBackend(nn.Module):
             verbose (bool): Enable verbose logging.
         """
         super().__init__()
-        w = str(model[0] if isinstance(model, list) else model)
         nn_module = isinstance(model, torch.nn.Module)
         (
             pt,
@@ -175,7 +176,7 @@ class AutoBackend(nn.Module):
             imx,
             rknn,
             triton,
-        ) = self._model_type(w)
+        ) = self._model_type("" if nn_module else model)
         fp16 &= pt or jit or onnx or xml or engine or nn_module or triton  # FP16
         nhwc = coreml or saved_model or pb or tflite or edgetpu or rknn  # BHWC formats (vs torch BCWH)
         stride, ch = 32, 3  # default stride and channels
@@ -189,8 +190,7 @@ class AutoBackend(nn.Module):
             cuda = False
 
         # Download if not local
-        if not (pt or triton or nn_module):
-            w = attempt_download_asset(w)
+        w = attempt_download_asset(model) if pt else model  # weights path
 
         # PyTorch (in-memory or file)
         if nn_module or pt:
@@ -203,11 +203,9 @@ class AutoBackend(nn.Module):
                     model = model.fuse(verbose=verbose)
                 model = model.to(device)
             else:  # pt file
-                from ultralytics.nn.tasks import attempt_load_weights
+                from ultralytics.nn.tasks import attempt_load_one_weight
 
-                model = attempt_load_weights(
-                    model if isinstance(model, list) else w, device=device, inplace=True, fuse=fuse
-                )
+                model, _ = attempt_load_one_weight(model, device=device, fuse=fuse)  # load model, ckpt
 
             # Common PyTorch model processing
             if hasattr(model, "kpt_shape"):
@@ -480,7 +478,7 @@ class AutoBackend(nn.Module):
 
         # TF.js
         elif tfjs:
-            raise NotImplementedError("YOLOv8 TF.js inference is not currently supported.")
+            raise NotImplementedError("Ultralytics TF.js inference is not currently supported.")
 
         # PaddlePaddle
         elif paddle:
@@ -612,9 +610,9 @@ class AutoBackend(nn.Module):
         im: torch.Tensor,
         augment: bool = False,
         visualize: bool = False,
-        embed: Optional[List] = None,
+        embed: list | None = None,
         **kwargs: Any,
-    ) -> Union[torch.Tensor, List[torch.Tensor]]:
+    ) -> torch.Tensor | list[torch.Tensor]:
         """
         Run inference on an AutoBackend model.
 
@@ -843,7 +841,7 @@ class AutoBackend(nn.Module):
         """
         return torch.tensor(x).to(self.device) if isinstance(x, np.ndarray) else x
 
-    def warmup(self, imgsz: Tuple[int, int, int, int] = (1, 3, 640, 640)) -> None:
+    def warmup(self, imgsz: tuple[int, int, int, int] = (1, 3, 640, 640)) -> None:
         """
         Warm up the model by running one forward pass with a dummy input.
 
@@ -857,7 +855,7 @@ class AutoBackend(nn.Module):
                 self.forward(im)  # warmup
 
     @staticmethod
-    def _model_type(p: str = "path/to/model.pt") -> List[bool]:
+    def _model_type(p: str = "path/to/model.pt") -> list[bool]:
         """
         Take a path to a model file and return the model type.