FifdZddlmZddlmZddlmZddlmZddl m Z ddl m Z Gd d eZ y ) ag Interface for Baidu's RT-DETR, a Vision Transformer-based real-time object detector. RT-DETR offers real-time performance and high accuracy, excelling in accelerated backends like CUDA with TensorRT. It features an efficient hybrid encoder and IoU-aware query selection for enhanced detection accuracy. References: https://arxiv.org/pdf/2304.08069.pdf )Model)RTDETRDetectionModel) TORCH_1_11)RTDETRPredictor) RTDETRTrainer)RTDETRValidatorcDeZdZdZddeddffd ZedefdZxZ S)RTDETRaInterface for Baidu's RT-DETR model, a Vision Transformer-based real-time object detector. This model provides real-time performance with high accuracy. It supports efficient hybrid encoding, IoU-aware query selection, and adaptable inference speed. Attributes: model (str): Path to the pre-trained model. Methods: task_map: Return a task map for RT-DETR, associating tasks with corresponding Ultralytics classes. Examples: Initialize RT-DETR with a pre-trained model >>> from ultralytics import RTDETR >>> model = RTDETR("rtdetr-l.pt") >>> results = model("image.jpg") modelreturnNcBtsJdt| |dy)zInitialize the RT-DETR model with the given pre-trained model file. Args: model (str): Path to the pre-trained model. Supports .pt, .yaml, and .yml formats. zRTDETR requires torch>=1.11detect)r taskN)rsuper__init__)selfr __class__s a/var/www/pod-logistic/pod-ai/venv/lib/python3.12/site-packages/ultralytics/models/rtdetr/model.pyrzRTDETR.__init__(s% 888z u84c4dttttdiS)zReturn a task map for RT-DETR, associating tasks with corresponding Ultralytics classes. Returns: (dict): A dictionary mapping task names to Ultralytics task classes for the RT-DETR model. r) predictor validatortrainerr )rr rr)rs rtask_mapzRTDETR.task_map1s" ,,(-   r)z rtdetr-l.pt) __name__ __module__ __qualname____doc__strrpropertydictr __classcell__)rs@rr r s6$5c5d5  $    rr N)rultralytics.engine.modelrultralytics.nn.tasksrultralytics.utils.torch_utilsrpredictrtrainrvalr r rrr+s*+54$ * U* r