how to turn a onnx model to pytorch model?

[lunasdejavu]

I want to use mobilenet-yolov2 so that I turned a pre-trained mobilenet model in caffe version to caffe2 version first . Then I followed the official tutorial to turn the caffe2 model to onnx model. then I can't find the method to turn onnx model to pytorch model since most of people turn pytorch to onnx now. Can someone help me? thanks

[prasanthpul]

Can you clarify what your end goal is? Do you want to inference the model? If so, you don't need to use PyTorch for that.

[lunasdejavu]

@prasanthpul I want to do object detection by combining mobilenet v1 and yolov2 so that I just need the mobilenet to do the feature extraction.

[askhade]

@lunasdejavu : What are you trying to achieve by converting the onnx model back to pytorch? If this onnx model? If this onnx model is the final model you want to run inference on then you can use onnxruntime : https://github.com/microsoft/onnxruntime