OronG13

Unfortunately no, Actually for now I left this issue with hope that when I will need it again it will already solved. I prefer to find out that there will...

Hello @grimoire, Thank you for your suggestion. We tried several values options, for example, we extremely reduced them to: pre_top_k=5, keep_top_k=5, max_output_boxes_per_class=1 But still, for every values combination we tested,...

Hello @grimoire, The original generated Onnx and PlugIn source code produce the value 6 for t_size variable. During our investigations, without full understanding the impacts, just for debugging, we manually...

We tried that test also. We manually set its value to 1. Still the CUDA error reproduced. Based on our tests that we made till now, I think that the...

Thank you @grimoire, Your support is appriciated. We tried to flat the NMS calls stack to one function without any real implementation changes but still got the error. Finally, we...