找到flow_from目录0张图片

时间:2020-04-21 09:27:52

标签: keras deep-learning

nb_classes = 5
lbls = list(map(str, range(nb_classes)))
batch_size = 32
img_size = 150
nb_epochs = 10
training_set = train_datagen.flow_from_dataframe(
    dataframe=df_train,
    directory='E:\\all\\DR\\train_images',
    x_col="id_code",
    y_col="diagnosis",
    batch_size=batch_size,
    shuffle=True,
    class_mode="categorical",
    classes=lbls,
    target_size=(img_size,img_size),
    subset='training')

validation_set = train_datagen.flow_from_dataframe(
    dataframe=df_train,
    directory='E:\\all\\DR\\train_images',
    x_col="id_code",
    y_col="diagnosis",
    batch_size=batch_size,
    shuffle=True,
    class_mode="categorical",
    classes=lbls,
    target_size=(img_size,img_size),
    subset='validation'
)

Found 0 validated image filenames belonging to 5 classes.
Found 0 validated image filenames belonging to 5 classes.

我试图将图像保存在单独的文件夹中,然后尝试访问,但是再次发生相同的问题,许多解决方案都无法解决问题

0 个答案:

没有答案