使用 colab 训练图像分类模型

Question

我按照模板更改了 link ，但它不起作用 https://colab.research.google.com/github/tensorflow/tensorflow/blob/master/tensorflow/lite/g3doc/tutorials/model_maker_image_classification.ipynb#scrollTo=3jz5x0JoskPv

这是我的数据集 https://firebasestorage.googleapis.com/v0/b/lol-fypproject.appspot.com/o/lol.tgz?alt=media&token=d07b81bd-442f-4ebe-920e-3772598fbb20

原码

image_path = tf.keras.utils.get_file(
  'flower_photos.tgz',
  'https://storage.googleapis.com/download.tensorflow.org/example_images/flower_photos.tgz',
  extract=True)
 image_path = os.path.join(os.path.dirname(image_path), 'flower_photos')

我改变了

image_path = tf.keras.utils.get_file(
  'lol.tgz',
  'https://firebasestorage.googleapis.com/v0/b/lol-fypproject.appspot.com/o/lol.tgz?alt=media&token=d07b81bd-442f-4ebe-920e-3772598fbb20',
  extract=True)
image_path = os.path.join(os.path.dirname(image_path), 'lol')

线路错误并显示错误信息

data = ImageClassifierDataLoader.from_folder(image_path)

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-15-a5e7646aca55> in <module>()
----> 1 data = ImageClassifierDataLoader.from_folder(image_path)
  2 train_data, test_data = data.split(0.9)

/usr/local/lib/python3.7/dist- 
packages/tensorflow_examples/lite/model_maker/core/data_util/image_dataloader.py 
 in 
 from_folder(cls, filename, shuffle)
 69     all_image_size = len(all_image_paths)
 70     if all_image_size == 0:
 ---> 71       raise ValueError('Image size is zero')
 72 
 73     if shuffle:

 ValueError: Image size is zero

Answer 1

我找到问题了

zip 文件的路径与示例的结构不正确

使用 colab 训练图像分类模型

train image classification models with colab

machine-learning

tensorflow

google-colaboratory

image-classification