ultralytics 8.3.118__tar.gz → 8.3.120__tar.gz

{ultralytics-8.3.118/ultralytics.egg-info → ultralytics-8.3.120}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ultralytics
-Version: 8.3.118
+Version: 8.3.120
 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>
@@ -32,7 +32,7 @@ Classifier: Operating System :: Microsoft :: Windows
 Requires-Python: >=3.8
 Description-Content-Type: text/markdown
 License-File: LICENSE
-Requires-Dist: numpy<=2.1.1,>=1.23.0
+Requires-Dist: numpy>=1.23.0
 Requires-Dist: matplotlib>=3.3.0
 Requires-Dist: opencv-python>=4.6.0
 Requires-Dist: pillow>=7.1.2
@@ -61,20 +61,19 @@ Requires-Dist: mkdocs-macros-plugin>=1.0.5; extra == "dev"
 Provides-Extra: export
 Requires-Dist: onnx>=1.12.0; extra == "export"
 Requires-Dist: coremltools>=8.0; (platform_system != "Windows" and python_version <= "3.13") and extra == "export"
-Requires-Dist: scikit-learn>=1.3.2; (platform_system != "Windows" and python_version <= "3.12") and extra == "export"
+Requires-Dist: scikit-learn>=1.3.2; (platform_system != "Windows" and python_version <= "3.13") and extra == "export"
 Requires-Dist: openvino>=2024.0.0; extra == "export"
 Requires-Dist: tensorflow>=2.0.0; extra == "export"
 Requires-Dist: tensorflowjs>=2.0.0; extra == "export"
 Requires-Dist: tensorstore>=0.1.63; (platform_machine == "aarch64" and python_version >= "3.9") and extra == "export"
-Requires-Dist: keras; extra == "export"
 Requires-Dist: h5py!=3.11.0; platform_machine == "aarch64" and extra == "export"
 Provides-Extra: solutions
 Requires-Dist: shapely<2.1.0,>=2.0.0; extra == "solutions"
 Requires-Dist: streamlit<1.44.0,>=1.29.0; extra == "solutions"
 Provides-Extra: logging
-Requires-Dist: comet; extra == "logging"
-Requires-Dist: tensorboard>=2.13.0; extra == "logging"
-Requires-Dist: dvclive>=2.12.0; extra == "logging"
+Requires-Dist: wandb; extra == "logging"
+Requires-Dist: tensorboard; extra == "logging"
+Requires-Dist: mlflow; extra == "logging"
 Provides-Extra: extra
 Requires-Dist: hub-sdk>=0.0.12; extra == "extra"
 Requires-Dist: ipython; extra == "extra"

{ultralytics-8.3.118 → ultralytics-8.3.120}/pyproject.toml

@@ -61,7 +61,7 @@ classifiers = [
 
 # Required dependencies ------------------------------------------------------------------------------------------------
 dependencies = [
-    "numpy>=1.23.0,<=2.1.1", # OpenVINO and TFLite errors on '--slow' CI Tests https://github.com/ultralytics/ultralytics/pull/18943
+    "numpy>=1.23.0",
     "matplotlib>=3.3.0",
     "opencv-python>=4.6.0",
     "pillow>=7.1.2",
@@ -95,12 +95,11 @@ dev = [
 export = [
     "onnx>=1.12.0", # ONNX export
     "coremltools>=8.0; platform_system != 'Windows' and python_version <= '3.13'", # CoreML supported on macOS and Linux
-    "scikit-learn>=1.3.2; platform_system != 'Windows' and python_version <= '3.12'", # CoreML k-means quantization
+    "scikit-learn>=1.3.2; platform_system != 'Windows' and python_version <= '3.13'", # CoreML k-means quantization
     "openvino>=2024.0.0",  # OpenVINO export
     "tensorflow>=2.0.0", # TF bug https://github.com/ultralytics/ultralytics/issues/5161
     "tensorflowjs>=2.0.0", # TF.js export, automatically installs tensorflow
     "tensorstore>=0.1.63; platform_machine == 'aarch64' and python_version >= '3.9'", # for TF Raspberry Pi exports
-    "keras", # not installed automatically by tensorflow>=2.16
     "h5py!=3.11.0; platform_machine == 'aarch64'", # fix h5py build issues due to missing aarch64 wheels in 3.11 release
 ]
 solutions = [
@@ -108,9 +107,9 @@ solutions = [
     "streamlit>=1.29.0,<1.44.0",    # for live inference on web browser, i.e `yolo streamlit-predict`
 ]
 logging = [
-    "comet", # https://docs.ultralytics.com/integrations/comet/
-    "tensorboard>=2.13.0",
-    "dvclive>=2.12.0",
+    "wandb", # https://docs.ultralytics.com/integrations/weights-biases/
+    "tensorboard",  # https://docs.ultralytics.com/integrations/tensorboard/
+    "mlflow",  # https://docs.ultralytics.com/integrations/mlflow/
 ]
 extra = [
     "hub-sdk>=0.0.12", # Ultralytics HUB

{ultralytics-8.3.118 → ultralytics-8.3.120}/ultralytics/__init__.py

@@ -1,6 +1,6 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
-__version__ = "8.3.118"
+__version__ = "8.3.120"
 
 import os
 

{ultralytics-8.3.118 → ultralytics-8.3.120}/ultralytics/cfg/__init__.py

@@ -181,6 +181,7 @@ CFG_FRACTION_KEYS = frozenset(
         "bgr",
         "mosaic",
         "mixup",
+        "cutmix",
         "copy_paste",
         "conf",
         "iou",

{ultralytics-8.3.118 → ultralytics-8.3.120}/ultralytics/cfg/default.yaml

@@ -114,6 +114,7 @@ fliplr: 0.5 # (float) image flip left-right (probability)
 bgr: 0.0 # (float) image channel BGR (probability)
 mosaic: 1.0 # (float) image mosaic (probability)
 mixup: 0.0 # (float) image mixup (probability)
+cutmix: 0.0 # (float) image cutmix (probability)
 copy_paste: 0.0 # (float) segment copy-paste (probability)
 copy_paste_mode: "flip" # (str) the method to do copy_paste augmentation (flip, mixup)
 auto_augment: randaugment # (str) auto augmentation policy for classification (randaugment, autoaugment, augmix)

{ultralytics-8.3.118 → ultralytics-8.3.120}/ultralytics/data/augment.py

@@ -317,7 +317,7 @@ class Compose:
 
 class BaseMixTransform:
     """
-    Base class for mix transformations like MixUp and Mosaic.
+    Base class for mix transformations like Cutmix, MixUp and Mosaic.
 
     This class provides a foundation for implementing mix transformations on datasets. It handles the
     probability-based application of transforms and manages the mixing of multiple images and labels.
@@ -348,7 +348,7 @@ class BaseMixTransform:
 
     def __init__(self, dataset, pre_transform=None, p=0.0) -> None:
         """
-        Initializes the BaseMixTransform object for mix transformations like MixUp and Mosaic.
+        Initializes the BaseMixTransform object for mix transformations like CutMix, MixUp and Mosaic.
 
         This class serves as a base for implementing mix transformations in image processing pipelines.
 
@@ -368,7 +368,7 @@ class BaseMixTransform:
 
     def __call__(self, labels):
         """
-        Applies pre-processing transforms and mixup/mosaic transforms to labels data.
+        Applies pre-processing transforms and cutmix/mixup/mosaic transforms to labels data.
 
         This method determines whether to apply the mix transform based on a probability factor. If applied, it
         selects additional images, applies pre-transforms if specified, and then performs the mix transform.
@@ -391,7 +391,7 @@ class BaseMixTransform:
         if isinstance(indexes, int):
             indexes = [indexes]
 
-        # Get images information will be used for Mosaic or MixUp
+        # Get images information will be used for Mosaic, CutMix or MixUp
         mix_labels = [self.dataset.get_image_and_label(i) for i in indexes]
 
         if self.pre_transform is not None:
@@ -401,16 +401,16 @@ class BaseMixTransform:
 
         # Update cls and texts
         labels = self._update_label_text(labels)
-        # Mosaic or MixUp
+        # Mosaic, CutMix or MixUp
         labels = self._mix_transform(labels)
         labels.pop("mix_labels", None)
         return labels
 
     def _mix_transform(self, labels):
         """
-        Applies MixUp or Mosaic augmentation to the label dictionary.
+        Applies CutMix, MixUp or Mosaic augmentation to the label dictionary.
 
-        This method should be implemented by subclasses to perform specific mix transformations like MixUp or
+        This method should be implemented by subclasses to perform specific mix transformations like CutMix, MixUp or
         Mosaic. It modifies the input label dictionary in-place with the augmented data.
 
         Args:
@@ -439,7 +439,7 @@ class BaseMixTransform:
             >>> indexes = transform.get_indexes()
             >>> print(indexes)  # [3, 18, 7, 2]
         """
-        raise NotImplementedError
+        return random.randint(0, len(self.dataset) - 1)
 
     @staticmethod
     def _update_label_text(labels):
@@ -877,7 +877,6 @@ class MixUp(BaseMixTransform):
         p (float): Probability of applying MixUp augmentation.
 
     Methods:
-        get_indexes: Returns a random index from the dataset.
         _mix_transform: Applies MixUp augmentation to the input labels.
 
     Examples:
@@ -906,24 +905,6 @@ class MixUp(BaseMixTransform):
         """
         super().__init__(dataset=dataset, pre_transform=pre_transform, p=p)
 
-    def get_indexes(self):
-        """
-        Get a random index from the dataset.
-
-        This method returns a single random index from the dataset, which is used to select an image for MixUp
-        augmentation.
-
-        Returns:
-            (int): A random integer index within the range of the dataset length.
-
-        Examples:
-            >>> mixup = MixUp(dataset)
-            >>> index = mixup.get_indexes()
-            >>> print(index)
-            42
-        """
-        return random.randint(0, len(self.dataset) - 1)
-
     def _mix_transform(self, labels):
         """
         Applies MixUp augmentation to the input labels.
@@ -949,6 +930,122 @@ class MixUp(BaseMixTransform):
         return labels
 
 
+class CutMix(BaseMixTransform):
+    """
+    Applies CutMix augmentation to image datasets as described in the paper https://arxiv.org/abs/1905.04899.
+
+    CutMix combines two images by replacing a random rectangular region of one image with the corresponding region from another image,
+    and adjusts the labels proportionally to the area of the mixed region.
+
+    Attributes:
+        dataset (Any): The dataset to which CutMix augmentation will be applied.
+        pre_transform (Callable | None): Optional transform to apply before CutMix.
+        p (float): Probability of applying CutMix augmentation.
+        beta (float): Beta distribution parameter for sampling the mixing ratio (default=1.0).
+        num_areas (int): Number of areas to try to cut and mix (default=3).
+
+    Methods:
+        _mix_transform: Applies CutMix augmentation to the input labels.
+        _rand_bbox: Generates random bounding box coordinates for the cut region.
+
+    Examples:
+        >>> from ultralytics.data.augment import CutMix
+        >>> dataset = YourDataset(...)  # Your image dataset
+        >>> cutmix = CutMix(dataset, p=0.5)
+        >>> augmented_labels = cutmix(original_labels)
+    """
+
+    def __init__(self, dataset, pre_transform=None, p=0.0, beta=1.0, num_areas=3) -> None:
+        """
+        Initializes the CutMix augmentation object.
+
+        Args:
+            dataset (Any): The dataset to which CutMix augmentation will be applied.
+            pre_transform (Callable | None): Optional transform to apply before CutMix.
+            p (float): Probability of applying CutMix augmentation.
+            beta (float): Beta distribution parameter for sampling the mixing ratio (default=1.0).
+            num_areas (int): Number of areas to try to cut and mix (default=3).
+        """
+        super().__init__(dataset=dataset, pre_transform=pre_transform, p=p)
+        self.beta = beta
+        self.num_areas = num_areas
+
+    def _rand_bbox(self, width, height):
+        """
+        Generates random bounding box coordinates for the cut region.
+
+        Args:
+            width (int): Width of the image.
+            height (int): Height of the image.
+
+        Returns:
+            (tuple): (x1, y1, x2, y2) coordinates of the bounding box.
+        """
+        # Sample mixing ratio from Beta distribution
+        lam = np.random.beta(self.beta, self.beta)
+
+        cut_ratio = np.sqrt(1.0 - lam)
+        cut_w = int(width * cut_ratio)
+        cut_h = int(height * cut_ratio)
+
+        # Random center
+        cx = np.random.randint(width)
+        cy = np.random.randint(height)
+
+        # Bounding box coordinates
+        x1 = np.clip(cx - cut_w // 2, 0, width)
+        y1 = np.clip(cy - cut_h // 2, 0, height)
+        x2 = np.clip(cx + cut_w // 2, 0, width)
+        y2 = np.clip(cy + cut_h // 2, 0, height)
+
+        return x1, y1, x2, y2
+
+    def _mix_transform(self, labels):
+        """
+        Applies CutMix augmentation to the input labels.
+
+        Args:
+            labels (dict): A dictionary containing the original image and label information.
+
+        Returns:
+            (dict): A dictionary containing the mixed image and adjusted labels.
+
+        Examples:
+            >>> cutter = CutMix(dataset)
+            >>> mixed_labels = cutter._mix_transform(labels)
+        """
+        # Get a random second image
+        h, w = labels["img"].shape[:2]
+
+        cut_areas = np.asarray([self._rand_bbox(w, h) for _ in range(self.num_areas)], dtype=np.float32)
+        ioa1 = bbox_ioa(cut_areas, labels["instances"].bboxes)  # (self.num_areas, num_boxes)
+        idx = np.nonzero(ioa1.sum(axis=1) <= 0)[0]
+        if len(idx) == 0:
+            return labels
+
+        labels2 = labels.pop("mix_labels")[0]
+        area = cut_areas[np.random.choice(idx)]  # randomle select one
+        ioa2 = bbox_ioa(area[None], labels2["instances"].bboxes).squeeze(0)
+        indexes2 = np.nonzero(ioa2 >= 0.01 if len(labels["instances"].segments) else 0.1)[0]
+
+        instances2 = labels2["instances"][indexes2]
+        instances2.convert_bbox("xyxy")
+        instances2.denormalize(w, h)
+
+        # Apply CutMix
+        x1, y1, x2, y2 = area.astype(np.int32)
+        labels["img"][y1:y2, x1:x2] = labels2["img"][y1:y2, x1:x2]
+
+        # Restrain instances2 to the random bounding border
+        instances2.add_padding(-x1, -y1)
+        instances2.clip(x2 - x1, y2 - y1)
+        instances2.add_padding(x1, y1)
+
+        labels["cls"] = np.concatenate([labels["cls"], labels2["cls"][indexes2]], axis=0)
+        labels["instances"] = Instances.concatenate([labels["instances"], instances2], axis=0)
+        return labels
+
+
 class RandomPerspective:
     """
     Implements random perspective and affine transformations on images and corresponding annotations.
@@ -1655,7 +1752,6 @@ class CopyPaste(BaseMixTransform):
         p (float): Probability of applying Copy-Paste augmentation.
 
     Methods:
-        get_indexes: Returns a random index from the dataset.
         _mix_transform: Applies Copy-Paste augmentation to the input labels.
         __call__: Applies the Copy-Paste transformation to images and annotations.
 
@@ -1672,10 +1768,6 @@ class CopyPaste(BaseMixTransform):
         assert mode in {"flip", "mixup"}, f"Expected `mode` to be `flip` or `mixup`, but got {mode}."
         self.mode = mode
 
-    def get_indexes(self):
-        """Returns a list of random indexes from the dataset for CopyPaste augmentation."""
-        return random.randint(0, len(self.dataset) - 1)
-
     def _mix_transform(self, labels):
         """Applies Copy-Paste augmentation to combine objects from another image into the current image."""
         labels2 = labels["mix_labels"][0]
@@ -2445,6 +2537,7 @@ def v8_transforms(dataset, imgsz, hyp, stretch=False):
         [
             pre_transform,
             MixUp(dataset, pre_transform=pre_transform, p=hyp.mixup),
+            CutMix(dataset, pre_transform=pre_transform, p=hyp.cutmix),
             Albumentations(p=1.0),
             RandomHSV(hgain=hyp.hsv_h, sgain=hyp.hsv_s, vgain=hyp.hsv_v),
             RandomFlip(direction="vertical", p=hyp.flipud),

{ultralytics-8.3.118 → ultralytics-8.3.120}/ultralytics/data/dataset.py

@@ -215,6 +215,7 @@ class YOLODataset(BaseDataset):
         if self.augment:
             hyp.mosaic = hyp.mosaic if self.augment and not self.rect else 0.0
             hyp.mixup = hyp.mixup if self.augment and not self.rect else 0.0
+            hyp.cutmix = hyp.cutmix if self.augment and not self.rect else 0.0
             transforms = v8_transforms(self, self.imgsz, hyp)
         else:
             transforms = Compose([LetterBox(new_shape=(self.imgsz, self.imgsz), scaleup=False)])
@@ -235,14 +236,15 @@ class YOLODataset(BaseDataset):
 
     def close_mosaic(self, hyp):
         """
-        Sets mosaic, copy_paste and mixup options to 0.0 and builds transformations.
+        Disable mosaic, copy_paste, mixup and cutmix augmentations by setting their probabilities to 0.0.
 
         Args:
             hyp (dict): Hyperparameters for transforms.
         """
-        hyp.mosaic = 0.0  # set mosaic ratio=0.0
-        hyp.copy_paste = 0.0  # keep the same behavior as previous v8 close-mosaic
-        hyp.mixup = 0.0  # keep the same behavior as previous v8 close-mosaic
+        hyp.mosaic = 0.0
+        hyp.copy_paste = 0.0
+        hyp.mixup = 0.0
+        hyp.cutmix = 0.0
         self.transforms = self.build_transforms(hyp)
 
     def update_labels_info(self, label):
@@ -441,7 +443,7 @@ class GroundingDataset(YOLODataset):
         """
         assert task in {"detect", "segment"}, "GroundingDataset currently only supports `detect` and `segment` tasks"
         self.json_file = json_file
-        super().__init__(*args, task=task, data={}, **kwargs)
+        super().__init__(*args, task=task, data={"channels": 3}, **kwargs)
 
     def get_img_files(self, img_path):
         """

{ultralytics-8.3.118 → ultralytics-8.3.120}/ultralytics/engine/exporter.py

@@ -95,7 +95,6 @@ from ultralytics.utils import (
     yaml_save,
 )
 from ultralytics.utils.checks import (
-    IS_PYTHON_MINIMUM_3_12,
     check_imgsz,
     check_is_path_safe,
     check_requirements,
@@ -549,7 +548,7 @@ class Exporter:
         """YOLO ONNX export."""
         requirements = ["onnx>=1.12.0"]
         if self.args.simplify:
-            requirements += ["onnxslim", "onnxruntime" + ("-gpu" if torch.cuda.is_available() else "")]
+            requirements += ["onnxslim>=0.1.46", "onnxruntime" + ("-gpu" if torch.cuda.is_available() else "")]
         check_requirements(requirements)
         import onnx  # noqa
 
@@ -569,12 +568,6 @@ class Exporter:
                 dynamic["output0"].pop(2)
         if self.args.nms and self.model.task == "obb":
             self.args.opset = opset_version  # for NMSModel
-            # OBB error https://github.com/pytorch/pytorch/issues/110859#issuecomment-1757841865
-            try:
-                torch.onnx.register_custom_op_symbolic("aten::lift_fresh", lambda g, x: x, opset_version)
-            except RuntimeError:  # it will fail if it's already registered
-                pass
-            check_requirements("onnxslim>=0.1.46")  # Older versions has bug with OBB
 
         with arange_patch(self.args):
             export_onnx(
@@ -650,7 +643,7 @@ class Exporter:
                 """Quantization transform function."""
                 data_item: torch.Tensor = data_item["img"] if isinstance(data_item, dict) else data_item
                 assert data_item.dtype == torch.uint8, "Input image must be uint8 for the quantization preprocessing"
-                im = data_item.numpy().astype(np.float32) / 255.0  # uint8 to fp16/32 and 0 - 255 to 0.0 - 1.0
+                im = data_item.numpy().astype(np.float32) / 255.0  # uint8 to fp16/32 and 0-255 to 0.0-1.0
                 return np.expand_dims(im, 0) if im.ndim == 3 else im
 
             # Generate calibration data for integer quantization
@@ -914,14 +907,13 @@ class Exporter:
             import tensorflow as tf  # noqa
         check_requirements(
             (
-                "keras",  # required by 'onnx2tf' package
                 "tf_keras",  # required by 'onnx2tf' package
                 "sng4onnx>=1.0.1",  # required by 'onnx2tf' package
                 "onnx_graphsurgeon>=0.3.26",  # required by 'onnx2tf' package
                 "ai-edge-litert>=1.2.0",  # required by 'onnx2tf' package
                 "onnx>=1.12.0",
                 "onnx2tf>=1.26.3",
-                "onnxslim>=0.1.31",
+                "onnxslim>=0.1.46",
                 "onnxruntime-gpu" if cuda else "onnxruntime",
                 "protobuf>=5",
             ),
@@ -1027,8 +1019,6 @@ class Exporter:
     @try_export
     def export_edgetpu(self, tflite_model="", prefix=colorstr("Edge TPU:")):
         """YOLO Edge TPU export https://coral.ai/docs/edgetpu/models-intro/."""
-        LOGGER.warning(f"{prefix} Edge TPU known bug https://github.com/ultralytics/ultralytics/issues/1185")
-
         cmd = "edgetpu_compiler --version"
         help_url = "https://coral.ai/docs/edgetpu/compiler/"
         assert LINUX, f"export only supported on Linux. See {help_url}"
@@ -1126,7 +1116,8 @@ class Exporter:
         """YOLO IMX export."""
         gptq = False
         assert LINUX, (
-            "export only supported on Linux. See https://developer.aitrios.sony-semicon.com/en/raspberrypi-ai-camera/documentation/imx500-converter"
+            "export only supported on Linux. "
+            "See https://developer.aitrios.sony-semicon.com/en/raspberrypi-ai-camera/documentation/imx500-converter"
         )
         if getattr(self.model, "end2end", False):
             raise ValueError("IMX export is not supported for end2end models.")
@@ -1274,81 +1265,12 @@ class Exporter:
 
         return f, None
 
-    def _add_tflite_metadata(self, file, use_flatbuffers=False):
+    def _add_tflite_metadata(self, file):
         """Add metadata to *.tflite models per https://ai.google.dev/edge/litert/models/metadata."""
-        if not use_flatbuffers:
-            import zipfile
-
-            with zipfile.ZipFile(file, "a", zipfile.ZIP_DEFLATED) as zf:
-                zf.writestr("metadata.json", json.dumps(self.metadata, indent=2))
-            return
+        import zipfile
 
-        if IS_PYTHON_MINIMUM_3_12:
-            LOGGER.warning(f"TFLite Support package may not be compatible with Python>=3.12 environments for {file}")
-
-        # Update old 'flatbuffers' included inside tensorflow package
-        check_requirements(("tflite_support", "flatbuffers>=23.5.26,<100; platform_machine == 'aarch64'"))
-        import flatbuffers
-
-        try:
-            # TFLite Support bug https://github.com/tensorflow/tflite-support/issues/954#issuecomment-2108570845
-            from tensorflow_lite_support.metadata import metadata_schema_py_generated as schema  # noqa
-            from tensorflow_lite_support.metadata.python import metadata  # noqa
-        except ImportError:  # ARM64 systems may not have the 'tensorflow_lite_support' package available
-            from tflite_support import metadata  # noqa
-            from tflite_support import metadata_schema_py_generated as schema  # noqa
-
-        # Create model info
-        model_meta = schema.ModelMetadataT()
-        model_meta.name = self.metadata["description"]
-        model_meta.version = self.metadata["version"]
-        model_meta.author = self.metadata["author"]
-        model_meta.license = self.metadata["license"]
-
-        # Label file
-        tmp_file = Path(file).parent / "temp_meta.txt"
-        with open(tmp_file, "w", encoding="utf-8") as f:
-            f.write(str(self.metadata))
-
-        label_file = schema.AssociatedFileT()
-        label_file.name = tmp_file.name
-        label_file.type = schema.AssociatedFileType.TENSOR_AXIS_LABELS
-
-        # Create input info
-        input_meta = schema.TensorMetadataT()
-        input_meta.name = "image"
-        input_meta.description = "Input image to be detected."
-        input_meta.content = schema.ContentT()
-        input_meta.content.contentProperties = schema.ImagePropertiesT()
-        input_meta.content.contentProperties.colorSpace = schema.ColorSpaceType.RGB
-        input_meta.content.contentPropertiesType = schema.ContentProperties.ImageProperties
-
-        # Create output info
-        output1 = schema.TensorMetadataT()
-        output1.name = "output"
-        output1.description = "Coordinates of detected objects, class labels, and confidence score"
-        output1.associatedFiles = [label_file]
-        if self.model.task == "segment":
-            output2 = schema.TensorMetadataT()
-            output2.name = "output"
-            output2.description = "Mask protos"
-            output2.associatedFiles = [label_file]
-
-        # Create subgraph info
-        subgraph = schema.SubGraphMetadataT()
-        subgraph.inputTensorMetadata = [input_meta]
-        subgraph.outputTensorMetadata = [output1, output2] if self.model.task == "segment" else [output1]
-        model_meta.subgraphMetadata = [subgraph]
-
-        b = flatbuffers.Builder(0)
-        b.Finish(model_meta.Pack(b), metadata.MetadataPopulator.METADATA_FILE_IDENTIFIER)
-        metadata_buf = b.Output()
-
-        populator = metadata.MetadataPopulator.with_model_file(str(file))
-        populator.load_metadata_buffer(metadata_buf)
-        populator.load_associated_files([str(tmp_file)])
-        populator.populate()
-        tmp_file.unlink()
+        with zipfile.ZipFile(file, "a", zipfile.ZIP_DEFLATED) as zf:
+            zf.writestr("metadata.json", json.dumps(self.metadata, indent=2))
 
     def _pipeline_coreml(self, model, weights_dir=None, prefix=colorstr("CoreML Pipeline:")):
         """YOLO CoreML pipeline."""

{ultralytics-8.3.118 → ultralytics-8.3.120}/ultralytics/engine/tuner.py

@@ -88,8 +88,9 @@ class Tuner:
             "flipud": (0.0, 1.0),  # image flip up-down (probability)
             "fliplr": (0.0, 1.0),  # image flip left-right (probability)
             "bgr": (0.0, 1.0),  # image channel bgr (probability)
-            "mosaic": (0.0, 1.0),  # image mixup (probability)
+            "mosaic": (0.0, 1.0),  # image mosaic (probability)
             "mixup": (0.0, 1.0),  # image mixup (probability)
+            "cutmix": (0.0, 1.0),  # image cutmix (probability)
             "copy_paste": (0.0, 1.0),  # segment copy-paste (probability)
         }
         self.args = get_cfg(overrides=args)

{ultralytics-8.3.118 → ultralytics-8.3.120}/ultralytics/models/rtdetr/val.py

@@ -63,6 +63,7 @@ class RTDETRDataset(YOLODataset):
         if self.augment:
             hyp.mosaic = hyp.mosaic if self.augment and not self.rect else 0.0
             hyp.mixup = hyp.mixup if self.augment and not self.rect else 0.0
+            hyp.cutmix = hyp.cutmix if self.augment and not self.rect else 0.0
             transforms = v8_transforms(self, self.imgsz, hyp, stretch=True)
         else:
             # transforms = Compose([LetterBox(new_shape=(self.imgsz, self.imgsz), auto=False, scale_fill=True)])

{ultralytics-8.3.118 → ultralytics-8.3.120}/ultralytics/models/yolo/world/train_world.py

@@ -3,7 +3,7 @@
 from ultralytics.data import YOLOConcatDataset, build_grounding, build_yolo_dataset
 from ultralytics.data.utils import check_det_dataset
 from ultralytics.models.yolo.world import WorldTrainer
-from ultralytics.utils import DEFAULT_CFG
+from ultralytics.utils import DEFAULT_CFG, LOGGER
 from ultralytics.utils.torch_utils import de_parallel
 
 
@@ -93,14 +93,14 @@ class WorldTrainerFromScratch(WorldTrainer):
         """
         gs = max(int(de_parallel(self.model).stride.max() if self.model else 0), 32)
         if mode != "train":
-            return build_yolo_dataset(self.args, img_path, batch, self.data, mode=mode, rect=mode == "val", stride=gs)
-        dataset = [
-            build_yolo_dataset(self.args, im_path, batch, self.data, stride=gs, multi_modal=True)
+            return build_yolo_dataset(self.args, img_path, batch, self.data, mode=mode, rect=False, stride=gs)
+        datasets = [
+            build_yolo_dataset(self.args, im_path, batch, self.training_data[im_path], stride=gs, multi_modal=True)
             if isinstance(im_path, str)
             else build_grounding(self.args, im_path["img_path"], im_path["json_file"], batch, stride=gs)
             for im_path in img_path
         ]
-        return YOLOConcatDataset(dataset) if len(dataset) > 1 else dataset[0]
+        return YOLOConcatDataset(datasets) if len(datasets) > 1 else datasets[0]
 
     def get_dataset(self):
         """
@@ -140,8 +140,20 @@ class WorldTrainerFromScratch(WorldTrainer):
         # NOTE: to make training work properly, set `nc` and `names`
         final_data["nc"] = data["val"][0]["nc"]
         final_data["names"] = data["val"][0]["names"]
+        # NOTE: add path with lvis path
+        final_data["path"] = data["val"][0]["path"]
         final_data["channels"] = data["val"][0]["channels"]
         self.data = final_data
+        if self.args.single_cls:  # consistent with base trainer
+            LOGGER.info("Overriding class names with single class.")
+            self.data["names"] = {0: "object"}
+            self.data["nc"] = 1
+        self.training_data = {}
+        for d in data["train"]:
+            if self.args.single_cls:
+                d["names"] = {0: "object"}
+                d["nc"] = 1
+            self.training_data[d["train"]] = d
         return final_data["train"], final_data["val"][0]
 
     def plot_training_labels(self):

{ultralytics-8.3.118 → ultralytics-8.3.120}/ultralytics/models/yolo/yoloe/__init__.py

@@ -1,7 +1,7 @@
 # Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
 
 from .predict import YOLOEVPDetectPredictor, YOLOEVPSegPredictor
-from .train import YOLOEPEFreeTrainer, YOLOEPETrainer, YOLOETrainer, YOLOEVPTrainer
+from .train import YOLOEPEFreeTrainer, YOLOEPETrainer, YOLOETrainer, YOLOETrainerFromScratch, YOLOEVPTrainer
 from .train_seg import YOLOEPESegTrainer, YOLOESegTrainer, YOLOESegTrainerFromScratch, YOLOESegVPTrainer
 from .val import YOLOEDetectValidator, YOLOESegValidator
 
@@ -18,4 +18,5 @@ __all__ = [
     "YOLOEPEFreeTrainer",
     "YOLOEVPDetectPredictor",
     "YOLOEVPSegPredictor",
+    "YOLOETrainerFromScratch",
 ]