Is there any better implementation available for yolov4 with Output layer name BatchedNMS, BatchedNMS_1, BatchedNMS_2?

I need a complete inference script for yolov4 models. Where they have implemented the BatchedNMS output layer implementation.
I tried myself, but I’m facing an issue with post-processing outputs, is there any open source resource available please comment below.

Yolov4 training done using Nvidia-tao-toolkit.