[onnxruntime] - session.run return negative number

[huyhoangle86]

Hi, I converted my Scaled YOLOv4 csp pytorch model which trained on custom datasets to Darknet using this function (https://github.com/WongKinYiu/ScaledYOLOv4/blob/yolov4-csp/models/models.py#L647) from that convert Darket weight to ONNX using this repo(https://github.com/linghu8812/tensorrt_inference/tree/master/Yolov4) Finally I got Scaled YOLOv4 csp ONNX version. Then I started inference the onnx model but when I execute session.run it gives me negative result as in the pic bellow which strongly affected performance of model causes detect bouding boxes to be incorrect. what do you think?

System information

• OS Platform and Distribution : Linux Ubuntu 20.4
• ONNX Runtime installed from (source or binary): Source
• ONNX Runtime version: 1.8
• Python version: 3.7

[huyhoangle86]

helppp

[faxu]

Can you share your converted ONNX model that you're running?

[stale[bot]]

This issue has been automatically marked as stale due to inactivity and will be closed in 7 days if no further activity occurs. If further support is needed, please provide an update and/or more details.