yolov8改进步骤

1.看视频:parse
2.修改fitness()函数
位置：ultralytics/utils/metrics.py 检索fitness(self)

def fitness(self):"""Model fitness as a weighted combination of metrics."""w = [0.0, 1.0, 0.0, 0.0]  # weights for [P, R, mAP@0.5, mAP@0.5:0.95]return (np.array(self.mean_results()) * w).sum()

作用：修改报错best.py的依据是百分百看recall召回率
3.创建dataset/data.yaml文件
把path改成data的绝对路径地址！！！

path: /public/home/test202306/zj/data
train: train
val: val
test:# Classes
names:0: debris on the front of the vehicle1: cover open2: layer detachment3: anti loosening wire breakage4: oil leakage

4.创建train.py

import warnings
warnings.filterwarnings('ignore')
from ultralytics import YOLOif __name__ == '__main__':str = 'yolov8-C2f-DCNv3.yaml'model = YOLO('ultralytics/cfg/models/Add/{}'.format(str))# model.load('yolov8n.pt') # 是否加载预训练权重,科研不建议大家加载否则很难提升精度model.train(data=r'dataset/data.yaml',cache=False,imgsz=640,epochs=200,single_cls=False,  # 是否是单类别检测batch=128,close_mosaic=10,workers=6,device='0',optimizer='SGD', # using SGD# resume='runs/train/exp21/weights/last.pt', # 如过想续训就设置last.pt的地址amp=False,  # 如果出现训练损失为Nan可以关闭ampproject='runs/train',name='exp_{}'.format(str[:-5]),  # 当前实验的名称)