RKNN-Toolkit2 inference segmentation fault/return None

[briliantnugraha]

Hi, I've tried to convert yolov10n.onnx to rank and it is exported successfully.

However, when running the rank model in the RK3588 orange pi 5, it will cause segmentation fault and auto exit issue. I've also tried to trim the model similarly to the yolov8 suggestion (before the post process), but it still returns "None".

What might be the issue btw?