train

[shuoshuoshuoer]

train

[dwddw]

def img2label_paths(img_paths):
    # Define label paths as a function of image paths
    sa = 'images'
    sb = 'labels'
    # sa, sb = os.sep + 'images' + os.sep, os.sep + 'labels' + os.sep  # /images/, /labels/ substring
    return [x.replace(sa, sb, 1).replace('_' + x.split('_')[-1], '.txt') for x in img_paths] #replace('.' + x.split('.')[-1], '.txt')

[icey-zhang]

可以看一下断点运行到这

SuperYOLO/utils/datasets.py

Line 407 in e6c5c72

self.label_files = img2label_paths(self.img_files) # labels

，确认一下图像和标签的路径对不对。

[ljs-1]

没有预训练权重的化怎么样训练自己的数据集成

[Joey-He]

您好，我这边在训练时，按照您的步骤先下载数据集然后transform,运行只训练RGB,还是出现标签找不到：No labels in D:\experienment\SuperYOLO-main\SuperYOLO-main\dataset\VEDAI_1024\images.cache. Can not train without labels；您知道这该如何解决么

您好 您解决了吗

[Joey-He]

def img2label_paths(img_paths):
    # Define label paths as a function of image paths
    sa = 'images'
    sb = 'labels'
    # sa, sb = os.sep + 'images' + os.sep, os.sep + 'labels' + os.sep  # /images/, /labels/ substring
    return [x.replace(sa, sb, 1).replace('_' + x.split('_')[-1], '.txt') for x in img_paths] #replace('.' + x.split('.')[-1], '.txt')

谢谢 成功运行了