ultralytics 8.3.216__tar.gz → 8.3.217__tar.gz

{ultralytics-8.3.216 → ultralytics-8.3.217}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ultralytics
-Version: 8.3.216
+Version: 8.3.217
 Summary: Ultralytics YOLO 🚀 for SOTA object detection, multi-object tracking, instance segmentation, pose estimation and image classification.
 Author-email: Glenn Jocher <glenn.jocher@ultralytics.com>, Jing Qiu <jing.qiu@ultralytics.com>
 Maintainer-email: Ultralytics <hello@ultralytics.com>

{ultralytics-8.3.216 → ultralytics-8.3.217}/ultralytics/__init__.py

@@ -1,6 +1,6 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
-__version__ = "8.3.216"
+__version__ = "8.3.217"
 
 import importlib
 import os

{ultralytics-8.3.216 → ultralytics-8.3.217}/ultralytics/data/build.py

@@ -181,7 +181,15 @@ def build_grounding(
     )
 
 
-def build_dataloader(dataset, batch: int, workers: int, shuffle: bool = True, rank: int = -1, drop_last: bool = False):
+def build_dataloader(
+    dataset,
+    batch: int,
+    workers: int,
+    shuffle: bool = True,
+    rank: int = -1,
+    drop_last: bool = False,
+    pin_memory: bool = True,
+):
     """
     Create and return an InfiniteDataLoader or DataLoader for training or validation.
 
@@ -192,6 +200,7 @@ def build_dataloader(dataset, batch: int, workers: int, shuffle: bool = True, ra
         shuffle (bool, optional): Whether to shuffle the dataset.
         rank (int, optional): Process rank in distributed training. -1 for single-GPU training.
         drop_last (bool, optional): Whether to drop the last incomplete batch.
+        pin_memory (bool, optional): Whether to use pinned memory for dataloader.
 
     Returns:
         (InfiniteDataLoader): A dataloader that can be used for training or validation.
@@ -214,7 +223,7 @@ def build_dataloader(dataset, batch: int, workers: int, shuffle: bool = True, ra
         num_workers=nw,
         sampler=sampler,
         prefetch_factor=4 if nw > 0 else None,  # increase over default 2
-        pin_memory=nd > 0,
+        pin_memory=nd > 0 and pin_memory,
         collate_fn=getattr(dataset, "collate_fn", None),
         worker_init_fn=seed_worker,
         generator=generator,

{ultralytics-8.3.216 → ultralytics-8.3.217}/ultralytics/engine/results.py

@@ -527,7 +527,7 @@ class Results(SimpleClass, DataExportMixin):
         """
         assert color_mode in {"instance", "class"}, f"Expected color_mode='instance' or 'class', not {color_mode}."
         if img is None and isinstance(self.orig_img, torch.Tensor):
-            img = (self.orig_img[0].detach().permute(1, 2, 0).contiguous() * 255).to(torch.uint8).cpu().numpy()
+            img = (self.orig_img[0].detach().permute(1, 2, 0).contiguous() * 255).byte().cpu().numpy()
 
         names = self.names
         is_obb = self.obb is not None

{ultralytics-8.3.216 → ultralytics-8.3.217}/ultralytics/models/yolo/detect/val.py

@@ -300,7 +300,13 @@ class DetectionValidator(BaseValidator):
         """
         dataset = self.build_dataset(dataset_path, batch=batch_size, mode="val")
         return build_dataloader(
-            dataset, batch_size, self.args.workers, shuffle=False, rank=-1, drop_last=self.args.compile
+            dataset,
+            batch_size,
+            self.args.workers,
+            shuffle=False,
+            rank=-1,
+            drop_last=self.args.compile,
+            pin_memory=self.training,
         )
 
     def plot_val_samples(self, batch: dict[str, Any], ni: int) -> None:

{ultralytics-8.3.216 → ultralytics-8.3.217}/ultralytics/models/yolo/segment/val.py

@@ -173,7 +173,7 @@ class SegmentationValidator(DetectionValidator):
         if gt_cls.shape[0] == 0 or preds["cls"].shape[0] == 0:
             tp_m = np.zeros((preds["cls"].shape[0], self.niou), dtype=bool)
         else:
-            iou = mask_iou(batch["masks"].flatten(1), preds["masks"].flatten(1))
+            iou = mask_iou(batch["masks"].flatten(1), preds["masks"].flatten(1).float())  # float, uint8
             tp_m = self.match_predictions(preds["cls"], gt_cls, iou).cpu().numpy()
         tp.update({"tp_m": tp_m})  # update tp with mask IoU
         return tp

{ultralytics-8.3.216 → ultralytics-8.3.217}/ultralytics/nn/autobackend.py

@@ -19,6 +19,7 @@ from PIL import Image
 from ultralytics.utils import ARM64, IS_JETSON, LINUX, LOGGER, PYTHON_VERSION, ROOT, YAML, is_jetson
 from ultralytics.utils.checks import check_requirements, check_suffix, check_version, check_yaml, is_rockchip
 from ultralytics.utils.downloads import attempt_download_asset, is_url
+from ultralytics.utils.nms import non_max_suppression
 
 
 def check_class_names(names: list | dict) -> dict[int, str]:
@@ -854,7 +855,10 @@ class AutoBackend(nn.Module):
         if any(warmup_types) and (self.device.type != "cpu" or self.triton):
             im = torch.empty(*imgsz, dtype=torch.half if self.fp16 else torch.float, device=self.device)  # input
             for _ in range(2 if self.jit else 1):
-                self.forward(im)  # warmup
+                self.forward(im)  # warmup model
+                warmup_boxes = torch.rand(1, 84, 16, device=self.device)  # 16 boxes works best empirically
+                warmup_boxes[:, :4] *= imgsz[-1]
+                non_max_suppression(warmup_boxes)  # warmup NMS
 
     @staticmethod
     def _model_type(p: str = "path/to/model.pt") -> list[bool]:

{ultralytics-8.3.216 → ultralytics-8.3.217}/ultralytics/utils/ops.py

@@ -557,7 +557,7 @@ def process_mask(protos, masks_in, bboxes, shape, upsample: bool = False):
     masks = crop_mask(masks, boxes=bboxes * ratios)  # CHW
     if upsample:
         masks = F.interpolate(masks[None], shape, mode="bilinear")[0]  # CHW
-    return masks.gt_(0.0)
+    return masks.gt_(0.0).byte()
 
 
 def process_mask_native(protos, masks_in, bboxes, shape):
@@ -577,7 +577,7 @@ def process_mask_native(protos, masks_in, bboxes, shape):
     masks = (masks_in @ protos.float().view(c, -1)).view(-1, mh, mw)
     masks = scale_masks(masks[None], shape)[0]  # CHW
     masks = crop_mask(masks, bboxes)  # CHW
-    return masks.gt_(0.0)
+    return masks.gt_(0.0).byte()
 
 
 def scale_masks(masks, shape, padding: bool = True):
@@ -674,7 +674,7 @@ def masks2segments(masks, strategy: str = "all"):
     from ultralytics.data.converter import merge_multi_segment
 
     segments = []
-    for x in masks.int().cpu().numpy().astype("uint8"):
+    for x in masks.byte().cpu().numpy():
         c = cv2.findContours(x, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)[0]
         if c:
             if strategy == "all":  # merge and concatenate all segments
@@ -701,7 +701,7 @@ def convert_torch2numpy_batch(batch: torch.Tensor) -> np.ndarray:
     Returns:
         (np.ndarray): Output NumPy array batch with shape (Batch, Height, Width, Channels) and dtype uint8.
     """
-    return (batch.permute(0, 2, 3, 1).contiguous() * 255).clamp(0, 255).to(torch.uint8).cpu().numpy()
+    return (batch.permute(0, 2, 3, 1).contiguous() * 255).clamp(0, 255).byte().cpu().numpy()
 
 
 def clean_str(s):

{ultralytics-8.3.216 → ultralytics-8.3.217}/ultralytics.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ultralytics
-Version: 8.3.216
+Version: 8.3.217
 Summary: Ultralytics YOLO 🚀 for SOTA object detection, multi-object tracking, instance segmentation, pose estimation and image classification.
 Author-email: Glenn Jocher <glenn.jocher@ultralytics.com>, Jing Qiu <jing.qiu@ultralytics.com>
 Maintainer-email: Ultralytics <hello@ultralytics.com>