// example for python user
OpenCV python version: 4.9.0
Operating System / Platform: Ubuntu 20.04; win 10
Python version: 3.8; 3.10
Detailed description
When I try to try to run the onnx file create by converting yolov9-e-converted.pt , it fail to predict on any image. After debug, I find the CBFuse layer output the wrong tensor, then every layer related to those layers will output error tensor.
Then I create a simplest onnx to reproduce it.
y = torch.sum(torch.stack([x, x]), axis=0)
so y = 2*x
System Information
// example for python user OpenCV python version: 4.9.0 Operating System / Platform: Ubuntu 20.04; win 10 Python version: 3.8; 3.10
Detailed description
When I try to try to run the onnx file create by converting yolov9-e-converted.pt , it fail to predict on any image. After debug, I find the
CBFuse
layer output the wrong tensor, then every layer related to those layers will output error tensor.Then I create a simplest onnx to reproduce it.
y = torch.sum(torch.stack([x, x]), axis=0)
soy = 2*x
Steps to reproduce
the onnx file: tmp0.zip
torch
opencv
output
Issue submission checklist