• Hardware Platform (Jetson / GPU)
Xavier
• DeepStream Version
6.0
• JetPack Version (valid for Jetson only)
4.6-b197
• TensorRT Version
8.0.1-1+cuda10.2
• NVIDIA GPU Driver Version (valid for GPU only)
10.2
I am trying to test maskrcnn in Xavier.
Enginer can’t be produced with the following errors.
ERROR: [TRT]: UffParser: Validator error: multilevel_propose_rois: Unsupported operation _MultilevelProposeROI_TRT
Then followed the instructions here and downloaded this version
6.0 GA 4.6 TRT 8.0.1 release/8.0
Configured with the following command.
cmake .. -DGPU_ARCHS="53 62 72 75" -DTRT_LIB_DIR=/usr/lib/aarch64-linux-gnu/ -DCMAKE_C_COMPILER=/usr/bin/gcc -DTRT_BIN_DIR=
pwd/out
When build the error is
/opt/nvidia/deepstream/deepstream-6.0/sources/TensorRT/plugin/efficientNMSPlugin/efficientNMSInference.cu:18:10: fatal error: cub/cub.cuh: No such file or directory
#include "cub/cub.cuh"
^~~~~~~~~~~~~
compilation terminated.
How to solve?