YOLOv8-LibTorch-CPP-Inference

YOLOv8 LibTorch Inference C++

This example demonstrates how to perform inference using YOLOv8 models in C++ with LibTorch API.

Dependencies

Dependency	Version
OpenCV	>=4.0.0
C++ Standard	>=17
Cmake	>=3.18
Libtorch	>=1.12.1

Usage

git clone ultralytics
cd ultralytics
pip install .
cd examples/YOLOv8-LibTorch-CPP-Inference

mkdir build
cd build
cmake ..
make
./yolov8_libtorch_inference

Exporting YOLOv8

To export YOLOv8 models:

yolo export model=yolov8s.pt imgsz=640 format=torchscript