/loco_yolov5

Yolov5 training repository for the LOCO dataset

Primary LanguageShell

Training repository for YOLOv5 and the LOCO dataset using the provided Docker image.

Requirements

  • Docker, Nvidia-Docker and Nvidia Driver, see here

Quickstart

  1. Spin up the container bash start.sh
  2. Optional: Within the container, install Weights&Biases: pip install wandb and login wandb login.
  3. Download the LOCO Dataset bash data/scripts/get_loco.sh
  4. Train a model on loco
$ python train.py --data loco_all.yaml  # train a model
$ python test.py --data loco_all.yaml --weights yolov5s.pt  # test a model for Precision, Recall and mAP
$ python detect.py --weights yolov5s.pt --source path/to/images  # run inference on images and videos