yolor

yolor PyTorch=>ONNX=>TensorRT

1.Reference

• yolor code: https://github.com/WongKinYiu/yolor
• yolor arxiv: YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors
• get yolor weights from here: yolor.pt yolorx.pt yolor-w6.pt yolor-e6.pt yolor-d6.pt yolor-e6e.pt

For more information, please refer this blog: https://blog.csdn.net/linghu8812/article/details/125741951?spm=1001.2014.3001.5501

2.Export ONNX Model

Use the following command to export onnx model: first download yolor models to folder weights,

git clone https://github.com/linghu8812/yolor.git
cd yolor
python export.py --weights ./weights/yolor.pt

if you want to export onnx model with 1280 image size add --img-size in command:

python export.py --weights ./weights/yolor-w6.pt --simplify --grid --img-size 1280

3.Build yolor_trt Project

cd ../  # in project directory
mkdir build && cd build
cmake ..
make -j

4.Run yolor_trt

• inference with yolor

cd ../../bin/
./tensorrt_inference yolor ../configs/yolor/config.yaml ../samples/detection_segmentation

5.Results: