You're reading an pre-release version of this documentation.
For the latest stable release version, please have a look at master.

Pytorch转ONNX示例

yolov5模型转ONNX

本示例中展示yolov5s.pt转ONNX的示例。

环境搭建

1mamba create -n conversion python==3.10
2mamba activate conversion
3git clone https://github.com/ultralytics/yolov5.git
4cd yolov5
5pip install -r requirements.txt

yolov5s模型转化示例

终端执行

1 python export.py --weights yolov5s.pt --include onnx

脚本会自动从官方地址下载相应的权重，执行完成终端会输出如下信息，yolov5s.onnx生成在当前目录下

1 ONNX: export success  19.8s, saved as yolov5s.onnx (28.0 MB)

yolov8模型转ONNX

本示例中展示yolov8s.pt转ONNX的示例

环境搭建

1mamba create -n conversion python==3.10
2mamba activate conversion
3git clone https://github.com/Pertical/YOLOv8.git
4cd YOLOv8
5pip install ultralytics

yolov8s模型转化示例

终端执行

1 yolo export model=yolov8s.pt format=onnx

脚本会自动从官方地址下载相应的权重，执行完成终端会输出如下信息，yolov8s.onnx生成在当前目录下

1 ONNX: export success  15.0s, saved as 'yolov8s.onnx' (42.8 MB)