Issue with onnx model

Hi, I used pgnet inference model and got nice results. Then I tried to convert it to onnx model using paddle2onnx and successfully converted. But I couldn’t produce the results using that onnx model. I’m facing the below issue:-

(1, 3, 300, 300) 
NodeArg(name='x', type='tensor(float)', shape=[None, 3, None, None])

---------------------------------------------------------------------------

RuntimeException                          Traceback (most recent call last)

<ipython-input-21-90e728a110ab> in <module>()
     17 print(ort_sess.get_inputs()[0])
     18 ort_inputs = {ort_sess.get_inputs()[0].name: x}
---> 19 ort_outs = ort_sess.run(None, ort_inputs)
     20 print(ort_outs)
     21 print("Exported model has been predict by ONNXRuntime!")

/usr/local/lib/python3.7/dist-packages/onnxruntime/capi/onnxruntime_inference_collection.py in run(self, output_names, input_feed, run_options)
    186             output_names = [output.name for output in self._outputs_meta]
    187         try:
--> 188             return self._sess.run(output_names, input_feed, run_options)
    189         except C.EPFail as err:
    190             if self._enable_fallback:

RuntimeException: [ONNXRuntimeError] : 6 : RUNTIME_EXCEPTION : Non-zero status code returned while running Add node. Name:'Add_30' Status Message: /onnxruntime_src/onnxruntime/core/providers/cpu/math/element_wise_ops.h:497 void onnxruntime::BroadcastIterator::Init(ptrdiff_t, ptrdiff_t) axis == 1 || axis == largest was false. Attempting to broadcast an axis by a dimension other than 1. 19 by 20

where (1, 3, 300, 300) is my input to the onnx model and [None, 3, None, None] is expected by the model. Means there is no issue with the input. Could you please share any view on this issue? Thanks.