Question

我是TensorFlow的新手，对Python的使用也不是很熟练。我正在学习以下教程：

如果我使用lambda定义输入函数（如本教程中所述），则一切正常：

def my_input_fn(features, targets, batch_size=1, shuffle=True, num_epochs=None) : 
    ...
    features, labels = ds.make_one_shot_iterator().get_next()
    return features, labels

_=linear_regressor.train(input_fn=lambda: my_input_fn(my_feature, targets), steps=100)

如果我按以下方式更改脚本：

def get_my_input_fn() :
    def my_input_func(features, targets, batch_size=1, shuffle=True, num_epochs=None) :
        ...
        features, labels = ds.make_one_shot_iterator().get_next()
        return features, labels
    return my_input_func

temp_my_input_fn=get_my_input_fn()
_=linear_regressor.train(input_fn=temp_my_input_fn(my_feature, targets), steps=100)

我收到一个例外：

Traceback (most recent call last):
  File "/usr/lib/python3.6/inspect.py", line 1126, in getfullargspec
    sigcls=Signature)
  File "/usr/lib/python3.6/inspect.py", line 2193, in _signature_from_callable
    raise TypeError('{!r} is not a callable object'.format(obj))
TypeError: ({'MeanHHInc': <tf.Tensor 'IteratorGetNext:0' shape=(?,) dtype=float64>}, <tf.Tensor 'IteratorGetNext:1' shape=(?,) dtype=int64>) is not a callable object

在两种情况下，my_input_function()接收相同的参数并返回相同的元组(<class 'dict'>, <class 'tensorflow.python.framework.ops.Tensor'>)（在调试器中看到）。

使用第二种方法时我该怎么办？

Answer 1

super().__init__()

将把计算函数的值分配给this.setState({ edit: snippet })，因此它不再是可调用的对象。

检查下面的示例以查看区别

input_fn=temp_my_input_fn(my_feature, targets)

Answer 2

下面的代码示例将解决此可调用错误：

def get_my_input_fn(features, targets, batch_size=1, shuffle=True, num_epochs=None) :
  def my_input_fn():
    """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
    """
    nonlocal features, targets, batch_size, shuffle, num_epochs

    # 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
  return my_input_fn

temp_my_input_fn=get_my_input_fn(my_feature, targets)
_ = linear_regressor.train(input_fn=temp_my_input_fn, steps=100)

LinearRegressor.train（）贯穿“ ...不是可调用对象”异常

2 个答案: