我需要一种方法来获取TensorFlow中任何类型的层(即Dense,Conv2D等)的输出张量的形状.根据文档,有output_shape属性可以解决此问题.但是,每次我访问它都会得到AttributedError.
import numpy as np
import tensorflow as tf
x = np.arange(0,8,dtype=np.float32).reshape((1,8))
x = tf.constant(value=x,dtype=tf.float32,verify_shape=True)
dense = tf.layers.Dense(units=2)
out = dense(x)
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
res = sess.run(fetches=out)
print(res)
print(dense.output_shape)
print(dense.output_shape)语句将产生错误消息：
AttributeError: The layer has never been called and thus has no defined output shape.
或print(dense.output)将产生：
AttributeError('Layer ' + self.name + ' has no inbound nodes.')
AttributeError: Layer dense_1 has no inbound nodes.
附言：
我知道在上面的示例中,我可以通过out.get_shape()获得输出张量的形状.但是我想知道为什么output_shape属性不起作用以及如何解决？
我该如何解决？定义输入层：
x = tf.keras.layers.Input(tensor=tf.ones(shape=(1,8)))
dense = tf.layers.Dense(units=2)
out = dense(x)
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
res = sess.run(fetches=out)
print(dense.output_shape) # shape = (1,2)
根据Keras documentation,如果层具有单个节点,则可以通过以下方式获取其输入张量,输出张量,输入形状和输出形状：
> layer.input
> layer.output
> layer.input_shape
> layer.output_shape
但是在上面的示例中,当我们调用layer.output_shape或其他属性时,它会引发看起来有些奇怪的异常.
如果深入到source code,则错误由入站节点引起.
if not self._inbound_nodes:
raise AttributeError('The layer has never been called '
'and thus has no defined output shape.')
这些inbound nodes是什么？
A Node describes the connectivity between two layers. Each time a layer is connected to some new input,
a node is added to layer._inbound_nodes.
Each time the output of a layer is used by another layer,
a node is added to layer._outbound_nodes.
如上所示,当self._inbounds_nodes为None时,它将引发异常.这意味着,当某个层未连接到输入层或更普遍时,先前的所有层都未连接到输入层,self._inbounds_nodes为空,这会导致问题.
请注意,示例中的x是张量而不是输入层.请参阅另一个示例以获取更多说明：
x = tf.keras.layers.Input(shape=(8,))
dense = tf.layers.Dense(units=2)
out = dense(x)
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
res = sess.run(fetches=out,Feed_dict={x: np.ones(shape=(1,8))})
print(res)
print(res.shape) # shape = (1,2)
print(dense.output_shape) # shape = (None,2)
很好,因为定义了输入层.
请注意,在您的示例中,out是张量. tf.shape()函数和.shape =(get_shape())之间的区别是：
tf.shape(x)returns a 1-D integer tensor representing the dynamic
shape of x. A dynamic shape will be known only at graph execution time.
x.shapereturns a Python tuple representing the static
shape of x. A static shape,known at graph definition time.
有关更多张量形状的信息,请访问：https://pgaleone.eu/tensorflow/2018/07/28/understanding-tensorflow-tensors-shape-static-dynamic/