I'm running yolov5 object detection in geforce RTX 2080 ti, installed CUDA, CUDNN, tensorflow-gpu, visual studio but still my program is not using GPU

I am running Yolov5 object detector in my workstation Nvidia GeForce RTX 2080 Ti. I followed all the procedures installing the necessary requirements: CUDA 10.1 CUDNN 7.6 tensorflow-gpu 2.2 Visual Studio Python 3.7.0

But still my machine is not using GPU enough, while 58% CPU is being utilized. I want it to utilize the maximum utilization of GPU. Can anyone please help how can I achieve this? Thanks alot.

I have installed: 
CUDA 10.1
CUDNN 7.6 
tensorflow-gpu 2.2 
Visual Studio Python 3.7.0

and tried running Yolov5 object detector but still my machine is not using GPU enough, while 58% CPU is being utilized. How do I make the maximum utilization of GPU. Can anyone please help how can I achieve this? Thanks alot.