ONNX to TensorRT Engine conversion issue

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

Hi,

I have converted my mxnet model to Onnx format,
Now wanted to do the infrencing using TensorRt. So used onnx-tensorrt project to do so but stuck at below error

==========================================

Input filename: resnet100.onnx
ONNX IR version: 0.0.3
Opset version: 8
Producer name:
Producer version:
Domain:
Model version: 0
Doc string:

Parsing model
While parsing node number 6 [PRelu → “relu0”]:
ERROR: /home/nano/Pallab/TRT/onnx-tensorrt/ModelImporter.cpp:143 In function importNode:
[8] No importer registered for op: PRelu

============================================

TensorRt version:- 5.1.6.1
Device:- Jetson Nano,

Going to the various forum I learned that Prelu op support is missing and can be added as a plugin.
What could be the best approach to do the same, So that TensorRT can be used for infrencing.

Regards
Pallab Sarkar

2

Hi,
Please reference the steps described here for extending TRT with custom layer using plugin APIs : https://docs.nvidia.com/deeplearning/sdk/tensorrt-developer-guide/index.html#extending

Thanks