Initializes an object that repeats a given sampler indefinitely.

Initializes the Exporter class. Args: cfg (str, optional): Path to a configuration file. Defaults to DEFAULT_CFG.

Initialize YOLOv8 object with temporal parameters, such as feature history, alpha and current features.

Iterates over the 'sampler' and yields its contents.

Return the length of the sources object.

Applies default class names to an input YAML file or returns numerical class names.

Clip YOLOv5 augmented inference tails.

Ensures that the experiment object is only created in a single process during distributed training.

Create metadata map for model predictions by groupings them based on image ID.

De-scale predictions following augmented inference (inverse operation).

Return sum of left and right DFL losses.

Train completed, evaluate and plot if specified by arguments.

Embeds box prompts.

Embeds mask inputs.

Embeds point prompts.

Join the ground truth and prediction annotations if they exist.

Return data.yaml file.

Format ground truth annotations for detection.

Format the image for YOLOv5 from Numpy array to PyTorch tensor.

Format YOLO predictions for object detection visualization.

convert polygon points to bitmap.

Create a YOLOv5 classification model from a YOLOv5 detection model.

Set YOLOv8 model configurations and define the model architecture.

Gets the batch size of the output given the batch size of the input prompts.

Return an experiment based on mode and project name.

Get auxiliary losses

Fetch and return model data from Ultralytics HUB.

Hardcoded function to handle deprecated config keys

Initializes a new model and infers the task type from the model head. Args: weights (str): model checkpoint to be loaded

Log plots to the NeptuneAI experiment logger.

Log an image as a plot in the plot section of ClearML. Args: title (str): The title of the plot. plot_path (str): The path t

Logs evaluation plots and label plots for the experiment.

Applies MixUp or Mosaic augmentation to the label dictionary.

This function takes a path to a model file and returns the model type Args: p: path to the model file. Defaults to path/

Create a 2x2 image mosaic.

Create a 3x3 image mosaic.

Initializes a new model and infers the task type from the model definitions. Args: cfg (str): model configuration file

YOLOv8 CoreML pipeline.

Perform augmentations on input image x and return augmented inference.

Return correct prediction matrix Arguments: detections (array[N, 6]), x1, y1, x2, y2, conf, class labels (arr

Recombine the separated attention heads into a single tensor.

Register signal handlers for SIGTERM and SIGINT signals to gracefully handle termination.

Reset all registered callbacks.

Reset arguments when loading a PyTorch model.

YOLOv8 resizes images during training and the label values are normalized based on this resized shape. This function rescales the bounding box

Initializes and sets the DistributedDataParallel parameters for training.

Builds dataloaders and optimizer on correct rank process.

Validate and format an image to numpy array.

Validate and format an image to torch.Tensor.

Begin a threaded heartbeat loop to report the agent's status to Ultralytics HUB.

Terminate the heartbeat loop.

Unzip data.zip.

Update labels.

Start a new tracklet.

Calculate decomposed Relative Positional Embeddings from :paper:`mvitv2`. https://github.com/facebookresearch/mvit/blob/19786631e330df9f3622e

Adjust bounding boxes to stick to image border if they are within a certain threshold. Args: boxes (torch.Tensor): (n, 4) im

Center.

All close FP32 vs AMP results.

Initialize.

Initialize.

Return the homography matrix based on the GCPs in the next line of the input GMC file.

Initialize.

Apply affine to bboxes only. Args: bboxes (ndarray): list of bboxes, xyxy format, with shape (num_bboxes, 4).

Applies image transformation to labels.

Applies transformations to input 'labels' and returns object instances.

Apply affine to keypoints. Args: keypoints (ndarray): keypoints, [N, 17, 3]. M (ndarray): affine matrix.

Apply affine to segments and generate new bboxes from segments. Args: segments (ndarray): list of segments, [num_samples

Applies semantic segmentation to an image.

Return box areas.

Attempt to fetch authentication via cookies and set id_token. User must be logged in to HUB and running in a supported browser.

Automatically estimate the best YOLO batch size to use a fraction of the available CUDA memory. Args: model (torch.nn.module): YOLO

Merges a list of source of different types into a list of numpy arrays or PIL images

Yield batches of data from the input arguments.

Transform bbox(xyxy) to dist(ltrb).