Dmitry Kurtaev

Please provide corresponding PyTorch code

Seems like I misunderstood you initially. The speed of the model if OK but on 4.8.0 it consumes more memory, is that right?

@likholat, @alexeyhorkin, do you still have an alternative solution with MO Python extension which allows replace MaxUpooling to known ops? Cannot find in https://github.com/openvinotoolkit/openvino_notebooks/pull/487 **Update**: Never mind @Rakshith2597, may I...

Seems fine, but can you try `operator_export_type=torch.onnx.OperatorExportTypes.ONNX_ATEN_FALLBACK`?

@Rakshith2597, got it, thanks! Is there a way to share model definition so I can reproduce and propose a solution?

@Rakshith2597, seems like the following steps helped me convert a model: 1. Restore normal unpooling usage: ```python self.unpool1 = nn.MaxUnpool2d(2, stride=2) ... unpool1 = torch.cat([self.unpool1(donv2, idxs1), conv1], 1) # not...

Not sure that ONNX Runtime can handle this layer. I've tested OpenVINO 2022.2 and inference was fine.

Hi, @DunguTmp! Will take a look at this parameters. For iGPU you might check https://github.com/dkurt/openvino_pytorch_layers/pull/16 by @ngaloppo. I will resolve merge conflict and push to main branch.

Please fix failed CUDA test: https://github.com/opencv/opencv/actions/runs/9075787020/job/24937114379?pr=25487

@CNOCycle, please open a PR to https://github.com/opencv/opencv_extra/ with the same branch name as here.