Python – tensorflow.IndexedSlices.dense\_shape 属性

TensorFlow是谷歌设计的开源Python库，用于开发机器学习模型和深度学习神经网络。

dense_shape返回一个包含相应密集张量形状的一维张量。

语法:
tensorflow.IndexedSlices.dense_shape

返回：它返回一个一维张量。

示例 1:

# Importing the library
import tensorflow as tf

# Initializing the input
data = tf.constant([1, 2, 3])

# Printing the input
print('data: ', data)

# Calculating result
res = tf.IndexedSlices(data, [0], 2)

# Finding dense shape
dense = res.dense_shape

# Printing the result
print('dense shape: ', dense)

输出:

data: tf.Tensor([1 2 3], shape=(3, ), dtype=int32)
dense shape: 2

示例 2:

# Importing the library
import tensorflow as tf

# Initializing the input
data = tf.constant([1, 2, 3])

# Printing the input
print('data: ', data)

# Calculating result
res = tf.IndexedSlices(data, [0])

# Finding dense shape
dense = res.dense_shape

# Printing the result
print('dense shape: ', dense)

输出:

data: tf.Tensor([1 2 3], shape=(3, ), dtype=int32)
dense shape: None