如何从谷歌机器学习速成课程中替换 make_one_shot_iterator ()

Question

我正在学习 Google 机器学习强化课程。但它使用的是 TensorFlow 1.x 版本，所以我打算更改练习以便能够在 TensorFlow 2.0 中运行它们。但我被困在那个练习中：

https://colab.research.google.com/notebooks/mlcc/first_steps_with_tensor_flow.ipynb?utm_source=mlcc&utm_campaign=colab-external&utm_medium=referral&utm_content=firststeps-colab&hl=es#scrollTo=7UwqGbbxP53O

具体代码：

    def my_input_fn(features, targets, batch_size=1, shuffle=True, num_epochs=None):
    """Trains a linear regression model of one feature.

    Args:
      features: pandas DataFrame of features
      targets: pandas DataFrame of targets
      batch_size: Size of batches to be passed to the model
      shuffle: True or False. Whether to shuffle the data.
      num_epochs: Number of epochs for which data should be repeated. None = repeat indefinitely
    Returns:
      Tuple of (features, labels) for next data batch
    """

    # Convert pandas data into a dict of np arrays.
    features = {key:np.array(value) for key,value in dict(features).items()}                                           

    # Construct a dataset, and configure batching/repeating.
    ds = Dataset.from_tensor_slices((features,targets)) # warning: 2GB limit
    ds = ds.batch(batch_size).repeat(num_epochs)

    # Shuffle the data, if specified.
    if shuffle:
      ds = ds.shuffle(buffer_size=10000)

    # Return the next batch of data.
    features, labels = ds.make_one_shot_iterator().get_next()
    return features, labels

我已将features, labels = ds.make_one_shot_iterator().get_next() 替换为features, labels = tf.compat.v1.data.make_one_shot_iterator(ds).get_next()

它似乎可以工作，但 make_one_shot_iterator() 已被弃用，那么，我该如何替换它？

也照https://github.com/tensorflow/tensorflow/issues/29252，我试过了

 features, labels = ds.__iter__()
    next(ds.__iter__())
    return features, labels

但它返回错误__iter __ () is only supported inside of tf.function or when eager execution is enabled.

我在 python 方面非常缺乏经验，并且作为业余爱好者遵循课程。关于如何解决它的任何想法？谢谢。

Answer 1

经过多次测试，python 挂起是本地问题。 为了替换features, labels = ds.make_one_shot_iterator (). Get_next ()，我尝试了几件事：

features, labels = ds.__iter__().get_next()

iterator = ds.__iter__()
features, labels = iterator.get_next()

it = iter(ds)
features, labels = next(it)

所有三个案例都返回__iter__() is only supported inside of tf.function or when eager execution is enabled. 所以我尝试了：

features, labels = ds
return ds

还有就是：

return features, labels

两者都返回相同的错误，最后我尝试了：

return ds

神秘地它起作用了，我不知道为什么，但它确实起作用了。