pytorch export onnx with dynamic batchsize not able to build tensorrt engine.

[ynma-hanvo]

i have done constant unfolding with the model first, the tensorrt-onnx-parse gives me the following error. graphShapeAnalyzer.cpp::analyzeShapes::1872] Error Code 4: Miscellaneous (IElementWiseLayer Mul_2509: broadcast dimensions must be conformable.

is it a tensorrt issue or model issue. thanks for your advice.