This repository provides a detailed guide for training YOLOv5 on custom datasets using PyTorch, with a specific focus on medical endoscopy object detection applications. It covers environment setup, dataset preparation, configuration files, and the training process. The project also references TensorRT acceleration for model inference.