Fatal error: NvInfer.h: No such file or directory

AI & Data Science Deep Learning (Training & Inference) TensorRT
, , ,
1

I’m trying to run setup.py from torch2trt folder for trt_pose model.
I’m on a Tesla V100 GPU azure cloud instance and I’m getting the error " fatal error: NvInfer.h: No such file or directory" when I try to run the setup.py file using command “python3 setup.py install --plugins”.
Can someone please help with what I am missing in the process.

Please note: I currently don’t have a jetson device and and have not installed jetpack jdk.

TensorRT Version: 8.5.3.1
GPU Type: Tesla V100
Nvidia Driver Version: 520.61.05
CUDA Version: 11.8
CUDNN Version: 8.9.0
Operating System + Version: Linux Ubuntu 18.04
Python Version (if applicable): 3.6

2

Refer following link:

And verify if you define a proper tensorrt path.

BTW, JetPack is for Jetson series products, not V100.

3

Moving to TensorRT forum for better support, thanks.

4

I’ve installed tensorrt using the tar file installation method and I’ve added the /lib path to LD_LIBRARY_PATH environment variable before running the setup file

I’ve followed the exact steps as mentioned in the below link for tar installation method

5

Hi.

For those who are facing the problem, please try out the below link. That is what worked for me.