Efficientnet_lite4 ONNX compilation failing

[cybla]

I am trying to compile the official ONNX version of Efficientnet_lite4 (from here) to run on the Jacinto7 but the compilation fails with the following error message:

onnxruntime.capi.onnxruntime_pybind11_state.RuntimeException: [ONNXRuntimeError] : 6 : RUNTIME_EXCEPTION : Non-zero status code returned while running Gemm node. Name:'gemm' Status Message: /home/a0230315/workarea/onnxrt/onnxruntime/onnxruntime/core/providers/cpu/math/gemm_helper.h:13 onnxruntime::GemmHelper::GemmHelper(const onnxruntime::TensorShape&, bool, const onnxruntime::TensorShape&, bool, const onnxruntime::TensorShape&) left.NumDimensions() == 2 || left.NumDimensions() == 1 was false. 

I suspect it comes from an unsuported Gemm layer according to the previous warning:

ALLOWLISTING : GEMM : Only supported Gemm Params are transA = 0, transB = 1, alpha = 1.0 and beta = 1.0. The same will processed as Inner product or fully connected layer in TIDL  -- file info - tidl_onnxImport.cpp , TIDL_onnxGemmBaseParams , 849 

Do you have an idea on how to solve this ?

[mash2612]

Hi @cybla , were you able to resolve this issue?