Hello!
I am trying to export yolo26n quantized, I tried with NMS free end2end=False, I tested and it works correctly with the post processing for the boxes.
When I try with the NMS-baked in (default) I am not getting correct predictions.
Is this a limitation of the nms during the quantization? or am I making a mistake?
