It's a personal backup repository, but I will try my best to help you understand how to use them.

Now it contains:

cd pytorch
python setup.py install --user(optioanl)
cd ../examples python
from dcn_v2_wrapper import DeformableConv2DLayer as DCN

1. Copy plugin folders from tensorrt to NVIDIA/TensorRT/plugin

2. Add relative head file and initializePlugin() to InferPlugin.cpp at proper place, for example

#include "dcnv2Plugin.h"
#include "yoloPlugin.h"

initializePlugin<nvinfer1::plugin::DCNv2PluginCreator>(logger, libNamespace); initializePlugin<nvinfer1::plugin::YoloPluginCreator>(logger, libNamespace);

3. Add name of plugin folder to PLUGIN_LISTS in CMakeLists.txt

4. Build and use libnvinfer_plugin.so following offical introduction.

There are two pytorch2tensorrt transfer scripts in examples to show how these plugins work.

cd examples
python mbv3_centernet_trt7.py

The evaluation output is as follow which are mean values of hm, wh, reg