Triton Inference server build from source

I’m trying to build triton inference server from source
1.I downloaded the git link
2.cd server-master
mkdir build_server
3.server-master/build_server$ cmake -DTRITON_ENABLE_TENSORRT=ON -DTRITON_ENABLE_CUSTOM=ON -DTRITON_ENABLE_ENSEMBLE=ON …/build
4.make server

I’m having error :
server-master/src/backends/tensorrt/autofill.cc:161:16: error: ‘class nvinfer1::ICudaEngine’ has no member named ‘hasImplicitBatchDimension’
if (engine_->hasImplicitBatchDimension())

Please help what to do next