**YOLO-LITE** is a real demo model developed to run on mobile devices such as a laptop or phone not connected to a graphics unit (GPU).The global model was trained on the **PASCAL VOC** dataset and then on the **COCO** dataset, achieving a mean organism detector (mAP) of 33.81% and 12.26%, respectively.So a model with just 7 layers would have to be 482 million FLOPS.