Description
I installed TensorRT and CUDNN on Windows for using with yoloV8. Following nvidia documentation (zip installation): TensorRT installation documentation
But when I ran this code on Python3.11:
import tensorrt
print(tensorrt.__version__)
assert tensorrt.Builder(tensorrt.Logger())
this happens: Output image
8.6.1
[11/26/2023-11:38:09] [TRT] [E] 6: [libLoader.cpp::nvinfer1::rt::DynamicLibrary::Impl::Impl::290] Error Code 6: Internal Error (Unable to load library: nvinfer_builder_resource.dll) Traceback (most recent call last): File "d:\Project\test-projects\ultra\main.py", line 3, in <module> assert tensorrt.Builder(tensorrt.Logger()) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ TypeError: pybind11::init(): factory function returned nullptr
"nvinfer_builder_resource.dll" this file is located on "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.1\bin" and this location is added on system environment path.
what can I do for resolve this error.
Environment
TensorRT Version: 8.6.1.6
GPU Type: NVIDIA GeForce RTX 4060
Nvidia Driver Version: 536.40
CUDA Version: 12.1
CUDNN Version: 8.9.0.131
Operating System + Version: Windows 11
Python Version (if applicable): 3.11
TensorFlow Version (if applicable):
PyTorch Version (if applicable): 2.1.1
Baremetal or Container (if container which image + tag):
Steps To Reproduce
It can be help you:
import torch
import tensorrt
print(tensorrt.__version__)
print(torch.version.cuda)
print(torch.backends.cudnn.version())
Output:
8.6.1
12.1
8801
Installation steps for TensorRT: https://docs.nvidia.com/deeplearning/tensorrt/archives/tensorrt-861/install-guide/index.html#installing-zip
Installation steps for CUDNN: https://docs.nvidia.com/deeplearning/cudnn/install-guide/index.html#install-windows
Unable to load library: nvinfer_builder_resource.dll resolve this error.