How to export Pytorch model with custom op to ONNX and run it in ONNX Runtime

[flower1s]

In this path of the project: https://github.com/onnx/tutorials/blob/main/PyTorchCustomOperator/README.md; I attempted to compile onnxruntime according to the specified tutorial, but the following error occurred.

[1] 1169287
-- Configuring done (0.3s)
-- Configuring done (0.3s)
-- Generating done (0.4s)
-- Generating done (0.4s)
-- Build files have been written to: /afs/ihep.ac.cn/users/m/miaofeng/miaofeng/runboss/python/onnx_work/pytorchcustomoperater/ort_custom_op/build
-- Build files have been written to: /afs/ihep.ac.cn/users/m/miaofeng/miaofeng/runboss/python/onnx_work/pytorchcustomoperater/ort_custom_op/build
[ 33%] Building CXX object CMakeFiles/customop.dir/custom_op.cc.o
/afs/ihep.ac.cn/users/m/miaofeng/miaofeng/runboss/python/onnx_work/pytorchcustomoperater/ort_custom_op/custom_op.cc:16:22: error: expected initializer before '<' token
16 | void GroupNormKernel::Compute(OrtKernelContext* context) {
| ^
make[2]: *** [CMakeFiles/customop.dir/build.make:76: CMakeFiles/customop.dir/custom_op.cc.o] Error 1
make[1]: *** [CMakeFiles/Makefile2:83: CMakeFiles/customop.dir/all] Error 2
make: *** [Makefile:91: all] Error 2
[1]+ Done cmake ..

I am unable to solve it personally and hope to receive detailed guidance.
Best wishes!