Transpose node at onnx

[AyaNasser96]

Hello,
Why are you search with transpose to get the output layer? (how did you know the transpose is the output)
because when I search with that keyword I found this:

• and I re-train the yolov4 with 1 class (so it's a custom model now), then converted it to onnx, this gives me a 2 output layers (boxes, confs), and I want to use that onnx with openvino, Do your code helps with that case? (because I found 2 only output layers at my model not 3 as yours)