# Ultralytics YOLOv3

A PyTorch implementation of the YOLOv3 real-time object detection model, with training, validation, inference and export to formats such as ONNX, TensorRT and CoreML.

Ultralytics YOLOv3 is a PyTorch implementation of the YOLOv3 (You Only Look Once, version 3) real-time object detection model. It frames detection as a single regression problem, predicting bounding boxes and class probabilities directly from a full image in one forward pass.

The repository packages three classic YOLOv3 variants, YOLOv3, YOLOv3-SPP and YOLOv3-tiny, with tooling for training, validation, inference and export, reusing shared utilities from the main ultralytics package. A model can be loaded through PyTorch Hub or run from the command line, which downloads weights automatically and saves results; a separate script checks accuracy, and an export script converts a trained model to TorchScript, ONNX, OpenVINO, TensorRT, CoreML or PaddlePaddle.

It requires Python 3.8 or later and PyTorch 1.8 or later, and is released under the AGPL-3.0 license, with a separate enterprise license available for commercial use.

Website: https://docs.ultralytics.com/models/yolov3/