BMW-YOLOv4-Training-Automation
BMW-YOLOv4-Training-Automation is a Linux-based framework built on AlexeyAB's darknet repository that enables no-code training of YOLOv4 and YOLOv3 deep learning object detection models. Users provide a labeled dataset or use the companion BMW-LabelTool-Lite for annotation, then launch training with minimal configuration via Docker containers supporting both GPU and CPU environments, including NVIDIA GPUs such as Volta, Xavier, and Turing. Training progress can be monitored through TensorBoard or a custom REST API with a GUI. The tool includes a sample dataset, configurable training parameters, automatic train/test splitting, and pre-configured YOLOv4-specific hyperparameters such as mosaic and blur. Designed for Ubuntu 18.04, it aims to make state-of-the-art deep learning model training accessible without requiring deep technical setup.