f7b9009c91
Signed-off-by: Glenn Jocher <glenn.jocher@ultralytics.com> Co-authored-by: UltralyticsAssistant <web@ultralytics.com> Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
35 lines
617 B
Markdown
35 lines
617 B
Markdown
# 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
|
|
|
|
```bash
|
|
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:
|
|
|
|
```bash
|
|
yolo export model=yolov8s.pt imgsz=640 format=torchscript
|
|
```
|