从头开始运行一个yolo11的分类模型

从头开始运行一个yolo11的分类模型

1.yolo11网址 https://docs.ultralytics.com/zh/models/yolo11/

2.找到模型并下载

3.下载ultralytics https://github.com/ultralytics/ultralytics

4.解压倒入文件夹

5.创建trian.py

from ultralytics import YOLO

import yaml
with open("/Users/lfzxmw/Downloads/ultralytics-main3/shuju.yaml", "r") as f:
    data = yaml.safe_load(f)
print(data)

if __name__ == '__main__':
    model = YOLO('/Users/lfzxmw/Downloads/ultralytics-main3/yolo11n-cls.pt')
    model.train(
        data='/Users/lfzxmw/Downloads/ultralytics-main3/shuju.yaml',
        imgsz=224,
        epochs=100,
        batch=16,
        device='mps',
        workers=10
    )

6.将下载好的yolo11n.pt拖入文件夹

7.创建shoji.yaml文件 确保训练集和验证集路径的正确

train: /Users/lfzxmw/Downloads/ultralytics-main3/classs/train
val: /Users/lfzxmw/Downloads/ultralytics-main3/classs/val
#test: /Users/lfzxmw/Downloads/class/test

nc: 5
names: ['1','2','3','4','5']

8.点击train.py就可以开始训练了