Update ultralytics requirement from >=8.4.65 to >=8.4.83 (#2437)

* Update ultralytics requirement from >=8.4.65 to >=8.4.83

Updates the requirements on [ultralytics](https://github.com/ultralytics/ultralytics) to permit the latest version.
- [Release notes](https://github.com/ultralytics/ultralytics/releases)
- [Commits](ultralytics/ultralytics@v8.4.65...v8.4.83)

---
updated-dependencies:
- dependency-name: ultralytics
  dependency-version: 8.4.83
  dependency-type: direct:production
...

Signed-off-by: dependabot[bot] <support@github.com>

* Auto-format by https://ultralytics.com/actions

---------

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: UltralyticsAssistant <web@ultralytics.com>

Author: dependabot[bot]

models/common.py

@@ -160,8 +160,7 @@ def __init__(self, c1, c2, shortcut=True, g=1, e=0.5):  # ch_in, ch_out, shortcu
         self.add = shortcut and c1 == c2
 
     def forward(self, x):
-        """Executes forward pass, performing convolutional ops and optional shortcut addition; expects input tensor x.
-        """
+        """Executes forward pass, performing convolutional ops and optional shortcut addition; expects input tensor x."""
         return x + self.cv2(self.cv1(x)) if self.add else self.cv2(self.cv1(x))
 
 
@@ -183,8 +182,7 @@ def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5):  # ch_in, ch_out, nu
         self.m = nn.Sequential(*(Bottleneck(c_, c_, shortcut, g, e=1.0) for _ in range(n)))
 
     def forward(self, x):
-        """Processes input through layers, combining outputs with activation and normalization for feature extraction.
-        """
+        """Processes input through layers, combining outputs with activation and normalization for feature extraction."""
         y1 = self.cv3(self.m(self.cv1(x)))
         y2 = self.cv2(x)
         return self.cv4(self.act(self.bn(torch.cat((y1, y2), 1))))
@@ -212,8 +210,7 @@ class C3(nn.Module):
     """Implements a CSP Bottleneck with 3 convolutions, optional shortcuts, group convolutions, and expansion factor."""
 
     def __init__(self, c1, c2, n=1, shortcut=True, g=1, e=0.5):  # ch_in, ch_out, number, shortcut, groups, expansion
-        """Initializes CSP Bottleneck with 3 convolutions, optional shortcuts, group convolutions, and expansion factor.
-        """
+        """Initializes CSP Bottleneck with 3 convolutions, optional shortcuts, group convolutions, and expansion factor."""
         super().__init__()
         c_ = int(c2 * e)  # hidden channels
         self.cv1 = Conv(c1, c_, 1, 1)

pyproject.toml

@@ -79,7 +79,7 @@ dependencies = [
     "pandas>=1.1.4",
     "packaging>=20.0", # general utilities
     "seaborn>=0.11.0", # plotting
-    "ultralytics>=8.4.65",
+    "ultralytics>=8.4.83",
 ]
 
 # Optional dependencies ------------------------------------------------------------------------------------------------

requirements.txt

@@ -17,7 +17,7 @@ thop>=0.1.1                  # Model profiling - FLOPs and parameter count
 torch>=1.8.0                 # Core PyTorch for training/inference
 torchvision>=0.9.0           # Torch utilities for vision (transforms, datasets)
 tqdm>=4.66.3                 # Progress bar in CLI
-ultralytics>=8.4.65          # YOLO framework library (models, training, utils)
+ultralytics>=8.4.83          # YOLO framework library (models, training, utils)
 
 # Logging ---------------------------------------------------------------------
 # tensorboard>=2.4.1         # Visual logging (scalars, images)

utils/activations.py

@@ -110,8 +110,7 @@ def __init__(self, c1, k=1, s=1, r=16):  # ch_in, kernel, stride, r
         # self.bn2 = nn.BatchNorm2d(c1)
 
     def forward(self, x):
-        """Apply the MetaAconC activation, generating beta from the input's spatial average via the bottleneck network.
-        """
+        """Apply the MetaAconC activation, generating beta from the input's spatial average via the bottleneck network."""
         y = x.mean(dim=2, keepdims=True).mean(dim=3, keepdims=True)
         # batch-size 1 bug/instabilities https://github.com/ultralytics/yolov5/issues/2891
         # beta = torch.sigmoid(self.bn2(self.fc2(self.bn1(self.fc1(y)))))  # bug/unstable

utils/augmentations.py

@@ -263,8 +263,7 @@ def cutout(im, labels, p=0.5):
 
 
 def mixup(im, labels, im2, labels2):
-    """Applies MixUp augmentation by blending images and labels; see https://arxiv.org/pdf/1710.09412.pdf for details.
-    """
+    """Applies MixUp augmentation by blending images and labels; see https://arxiv.org/pdf/1710.09412.pdf for details."""
     r = np.random.beta(32.0, 32.0)  # mixup ratio, alpha=beta=32.0
     im = (im * r + im2 * (1 - r)).astype(np.uint8)
     labels = np.concatenate((labels, labels2), 0)

utils/dataloaders.py

@@ -481,8 +481,7 @@ def __len__(self):
 
 
 def img2label_paths(img_paths):
-    """Converts image paths to corresponding label paths by replacing `/images/` with `/labels/` and `.jpg` with `.txt`.
-    """
+    """Converts image paths to corresponding label paths by replacing `/images/` with `/labels/` and `.jpg` with `.txt`."""
     sa, sb = f"{os.sep}images{os.sep}", f"{os.sep}labels{os.sep}"  # /images/, /labels/ substrings
     return [sb.join(x.rsplit(sa, 1)).rsplit(".", 1)[0] + ".txt" for x in img_paths]
 

utils/general.py

@@ -197,8 +197,7 @@ class Profile(contextlib.ContextDecorator):
     """Profiles code execution time, usable as a context manager or decorator for performance monitoring."""
 
     def __init__(self, t=0.0):
-        """Initializes a profiling context for YOLOv3 with optional timing threshold `t` and checks CUDA availability.
-        """
+        """Initializes a profiling context for YOLOv3 with optional timing threshold `t` and checks CUDA availability."""
         self.t = t
         self.cuda = torch.cuda.is_available()
 
@@ -225,8 +224,7 @@ class Timeout(contextlib.ContextDecorator):
     """Enforces a timeout on code execution, raising TimeoutError on expiry."""
 
     def __init__(self, seconds, *, timeout_msg="", suppress_timeout_errors=True):
-        """Initializes a timeout context/decorator with specified duration, custom message, and error handling option.
-        """
+        """Initializes a timeout context/decorator with specified duration, custom message, and error handling option."""
         self.seconds = int(seconds)
         self.timeout_message = timeout_msg
         self.suppress = bool(suppress_timeout_errors)
@@ -311,15 +309,13 @@ def get_default_args(func):
 
 
 def get_latest_run(search_dir="."):
-    """Returns path to the most recent 'last.pt' file within 'search_dir' for resuming, or an empty string if not found.
-    """
+    """Returns path to the most recent 'last.pt' file within 'search_dir' for resuming, or an empty string if not found."""
     last_list = glob.glob(f"{search_dir}/**/last*.pt", recursive=True)
     return max(last_list, key=os.path.getctime) if last_list else ""
 
 
 def check_online():
-    """Checks internet connectivity by attempting to connect to "1.1.1.1" on port 443 twice; returns True if successful.
-    """
+    """Checks internet connectivity by attempting to connect to "1.1.1.1" on port 443 twice; returns True if successful."""
     import socket
 
     def run_once():
@@ -614,12 +610,10 @@ def url2file(url):
 
 
 def download(url, dir=".", unzip=True, delete=True, curl=False, threads=1, retry=3):
-    """Downloads files from URLs into a specified directory, optionally unzips, and supports multithreading and retries.
-    """
+    """Downloads files from URLs into a specified directory, optionally unzips, and supports multithreading and retries."""
 
     def download_one(url, dir):
-        """Downloads a file from a URL into the specified directory, supporting retries and using curl or torch methods.
-        """
+        """Downloads a file from a URL into the specified directory, supporting retries and using curl or torch methods."""
         success = True
         if os.path.isfile(url):
             f = Path(url)  # filename

utils/loss.py

@@ -83,8 +83,7 @@ def __init__(self, loss_fcn, gamma=1.5, alpha=0.25):
         self.loss_fcn.reduction = "none"  # required to apply FL to each element
 
     def forward(self, pred, true):
-        """Computes focal loss between predictions and true labels using configured loss function, `gamma`, and `alpha`.
-        """
+        """Computes focal loss between predictions and true labels using configured loss function, `gamma`, and `alpha`."""
         loss = self.loss_fcn(pred, true)
 
         pred_prob = torch.sigmoid(pred)  # prob from logits
@@ -187,8 +186,7 @@ def __call__(self, p, targets):  # predictions, targets
         return (lbox + lobj + lcls) * bs, torch.cat((lbox, lobj, lcls)).detach()
 
     def build_targets(self, p, targets):
-        """Match `targets` (image, class, x, y, w, h) to anchors per layer, returning tcls, tbox, indices, and anchors.
-        """
+        """Match `targets` (image, class, x, y, w, h) to anchors per layer, returning tcls, tbox, indices, and anchors."""
         na, nt = self.na, targets.shape[0]  # number of anchors, targets
         tcls, tbox, indices, anch = [], [], [], []
         gain = torch.ones(7, device=self.device)  # normalized to gridspace gain

utils/torch_utils.py

@@ -184,8 +184,7 @@ def de_parallel(model):
 
 
 def sparsity(model):
-    """Calculates and returns the global sparsity of a model as the ratio of zero-valued parameters to total parameters.
-    """
+    """Calculates and returns the global sparsity of a model as the ratio of zero-valued parameters to total parameters."""
     a, b = 0, 0
     for p in model.parameters():
         a += p.numel()

utils/triton.py

@@ -71,8 +71,7 @@ def __call__(self, *args, **kwargs) -> torch.Tensor | tuple[torch.Tensor, ...]:
         return result[0] if len(result) == 1 else result
 
     def _create_inputs(self, *args, **kwargs):
-        """Generates model inputs from args or kwargs, not allowing both; raises error if neither or both are provided.
-        """
+        """Generates model inputs from args or kwargs, not allowing both; raises error if neither or both are provided."""
         args_len, kwargs_len = len(args), len(kwargs)
         if not args_len and not kwargs_len:
             raise RuntimeError("No inputs provided.")