dgenerate-ultralytics-headless 8.3.157__tar.gz → 8.3.158__tar.gz

{dgenerate_ultralytics_headless-8.3.157 → dgenerate_ultralytics_headless-8.3.158}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dgenerate-ultralytics-headless
-Version: 8.3.157
+Version: 8.3.158
 Summary: Automatically built Ultralytics package with python-opencv-headless dependency instead of python-opencv
 Author-email: Glenn Jocher <glenn.jocher@ultralytics.com>, Jing Qiu <jing.qiu@ultralytics.com>
 Maintainer-email: Ultralytics <hello@ultralytics.com>

{dgenerate_ultralytics_headless-8.3.157 → dgenerate_ultralytics_headless-8.3.158}/dgenerate_ultralytics_headless.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dgenerate-ultralytics-headless
-Version: 8.3.157
+Version: 8.3.158
 Summary: Automatically built Ultralytics package with python-opencv-headless dependency instead of python-opencv
 Author-email: Glenn Jocher <glenn.jocher@ultralytics.com>, Jing Qiu <jing.qiu@ultralytics.com>
 Maintainer-email: Ultralytics <hello@ultralytics.com>

{dgenerate_ultralytics_headless-8.3.157 → dgenerate_ultralytics_headless-8.3.158}/ultralytics/__init__.py

@@ -1,6 +1,6 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
-__version__ = "8.3.157"
+__version__ = "8.3.158"
 
 import os
 

{dgenerate_ultralytics_headless-8.3.157 → dgenerate_ultralytics_headless-8.3.158}/ultralytics/data/augment.py

@@ -2928,7 +2928,7 @@ class ToTensor:
         the color channels are reversed from BGR to RGB.
 
         Args:
-            im (numpy.ndarray): Input image as a numpy array with shape (H, W, C) in BGR order.
+            im (numpy.ndarray): Input image as a numpy array with shape (H, W, C) in RGB order.
 
         Returns:
             (torch.Tensor): The transformed image as a PyTorch tensor in float32 or float16, normalized
@@ -2941,7 +2941,7 @@ class ToTensor:
             >>> print(tensor_img.shape, tensor_img.dtype)
             torch.Size([3, 640, 640]) torch.float16
         """
-        im = np.ascontiguousarray(im.transpose((2, 0, 1))[::-1])  # HWC to CHW -> BGR to RGB -> contiguous
+        im = np.ascontiguousarray(im.transpose((2, 0, 1)))  # HWC to CHW -> contiguous
         im = torch.from_numpy(im)  # to torch
         im = im.half() if self.half else im.float()  # uint8 to fp16/32
         im /= 255.0  # 0-255 to 0.0-1.0

{dgenerate_ultralytics_headless-8.3.157 → dgenerate_ultralytics_headless-8.3.158}/ultralytics/models/yolo/classify/predict.py

@@ -19,7 +19,6 @@ class ClassificationPredictor(BasePredictor):
 
     Attributes:
         args (dict): Configuration arguments for the predictor.
-        _legacy_transform_name (str): Name of the legacy transform class for backward compatibility.
 
     Methods:
         preprocess: Convert input images to model-compatible format.
@@ -50,7 +49,6 @@ class ClassificationPredictor(BasePredictor):
         """
         super().__init__(cfg, overrides, _callbacks)
         self.args.task = "classify"
-        self._legacy_transform_name = "ultralytics.yolo.data.augment.ToTensor"
 
     def setup_source(self, source):
         """Set up source and inference mode and classify transforms."""
@@ -58,22 +56,18 @@ class ClassificationPredictor(BasePredictor):
         updated = (
             self.model.model.transforms.transforms[0].size != max(self.imgsz)
             if hasattr(self.model.model, "transforms") and hasattr(self.model.model.transforms.transforms[0], "size")
-            else True
+            else False
+        )
+        self.transforms = (
+            classify_transforms(self.imgsz) if updated or not self.model.pt else self.model.model.transforms
         )
-        self.transforms = self.model.model.transforms if not updated else classify_transforms(self.imgsz)
 
     def preprocess(self, img):
         """Convert input images to model-compatible tensor format with appropriate normalization."""
         if not isinstance(img, torch.Tensor):
-            is_legacy_transform = any(
-                self._legacy_transform_name in str(transform) for transform in self.transforms.transforms
+            img = torch.stack(
+                [self.transforms(Image.fromarray(cv2.cvtColor(im, cv2.COLOR_BGR2RGB))) for im in img], dim=0
             )
-            if is_legacy_transform:  # Handle legacy transforms
-                img = torch.stack([self.transforms(im) for im in img], dim=0)
-            else:
-                img = torch.stack(
-                    [self.transforms(Image.fromarray(cv2.cvtColor(im, cv2.COLOR_BGR2RGB))) for im in img], dim=0
-                )
         img = (img if isinstance(img, torch.Tensor) else torch.from_numpy(img)).to(self.model.device)
         return img.half() if self.model.fp16 else img.float()  # Convert uint8 to fp16/32
 

{dgenerate_ultralytics_headless-8.3.157 → dgenerate_ultralytics_headless-8.3.158}/ultralytics/solutions/streamlit_inference.py

@@ -4,12 +4,15 @@ import io
 from typing import Any, List
 
 import cv2
+import torch
 
 from ultralytics import YOLO
 from ultralytics.utils import LOGGER
 from ultralytics.utils.checks import check_requirements
 from ultralytics.utils.downloads import GITHUB_ASSETS_STEMS
 
+torch.classes.__path__ = []  # Torch module __path__._path issue: https://github.com/datalab-to/marker/issues/442
+
 
 class Inference:
     """
@@ -133,7 +136,15 @@ class Inference:
     def configure(self) -> None:
         """Configure the model and load selected classes for inference."""
         # Add dropdown menu for model selection
-        available_models = [x.replace("yolo", "YOLO") for x in GITHUB_ASSETS_STEMS if x.startswith("yolo11")]
+        M_ORD, T_ORD = ["yolo11n", "yolo11s", "yolo11m", "yolo11l", "yolo11x"], ["", "-seg", "-pose", "-obb", "-cls"]
+        available_models = sorted(
+            [
+                x.replace("yolo", "YOLO")
+                for x in GITHUB_ASSETS_STEMS
+                if any(x.startswith(b) for b in M_ORD) and "grayscale" not in x
+            ],
+            key=lambda x: (M_ORD.index(x[:7].lower()), T_ORD.index(x[7:].lower() or "")),
+        )
         if self.model_path:  # If user provided the custom model, insert model without suffix as *.pt is added later
             available_models.insert(0, self.model_path.split(".pt", 1)[0])
         selected_model = self.st.sidebar.selectbox("Model", available_models)