Error in converting onnx model to IR

[TarunM]

I am getting errors in covering the attached onnx model (model.7z) to openvino.

(The model works perfectly with onnx, I have tested).

log.txt is the error output from the model optimizer.

-----------------------------UPDTE-------------------------------

After searching I got to know that is due to unsupported opcode=13.

So I converted the model directly from the tensorflow code.

But now I am getting following error in loading the to Inference Engine, please Help:-

exec_net = ie.load_network(network=net, device_name=device)
File "ie_api.pyx", line 336, in openvino.inference_engine.ie_api.IECore.load_network
File "ie_api.pyx", line 345, in openvino.inference_engine.ie_api.IECore.load_network
RuntimeError: Check 'element::Type::merge(inputs_et, inputs_et, get_input_element_type(i))' failed at C:\j\workspace\private-ci\ie\build-windows-vs2019@2\b\repos\openvino\ngraph\core\src\op\concat.cpp:62:
While validating node 'v0::Concat StatefulPartitionedCall/functional_1/tf_op_layer_Shape_9/Shape_9 (DynDim/Gather_5800[0]:i64{1}, ConstDim/Constant_5801[0]:i32{1}) -> (i32{2})' with friendly_name 'StatefulPartitionedCall/functional_1/tf_op_layer_Shape_9/Shape_9':
Argument element types are inconsistent.

(Attached is the converted xml, IR is not uploaded here, let me know if required)

[IntelSupport]

Hi TarunM,

Thanks for reaching out.Which onnx model you are using?What is your OpenVINO version? Meanwhile, you are attaching the .xml file which is the IR file that generated from Model Optimizer of OpenVINO. If you are getting the IR file then its mean your optimization is successful.

Regarding the "loading to Inference Engine", could you share your workaround and command for us to have a clear understanding.

Regards,

Aznie

[TarunM]

I dont know, I was able to convert IR but it was failing to load. Due to so many difficulties, I have abandoned this work and I do not have time to work on this one. sorry.

[IntelSupport]

Hi TarunM,

Since you don’t want to proceed for further support we will close the ticket. Thank you for your question. If you need any additional information from Intel, please submit a new question as this thread is no longer being monitored. 

Regards,

Aznie