Can I use PyTorch models with the VPU plugin?

[Roxy1]

I am trying to run my deep learning models on the Intel VPU using the OpenVINO toolkit. According to the Supported Networks for VPU devices on this page:

https://docs.openvino.ai/latest/openvino_docs_OV_UG_supported_plugins_VPU.html#doxid-openvino-docs-o...

, PyTorch is not listed as a supported network for the VPU plugin. Does this mean that I cannot use PyTorch models with the VPU plugin?

Thank you for your help.

[Peh_Intel]

Hi Roxy1,

Thanks for reaching out to us.

As long as the layers in PyTorch model is supported (listed in Supported Layers), the model can be used with VPU plugin.

You have to convert PyTorch model into ONNX first. Next, you can directly load ONNX model into VPU plugin or convert ONNX model to Intermediate Representation (IR) and then load IR into VPU plugin.

Regards,

Peh

[Peh_Intel]

Hi Roxy1,

This thread will no longer be monitored since we have provided answer. If you need any additional information from Intel, please submit a new question. 

Regards,

Peh