TensorFlow – 如何从张量创建一个 numpy ndarray

TensorFlow 是由 Google 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。

为了从张量创建一个 numpy 数组，张量首先被转换为一个原型张量。

Method Used:

• make_ndarray: This method accepts a TensorProto as input and returns a numpy array with same content as TensorProto.

示例 1：

Python3

# importing the library
import tensorflow as tf
  
# Initializing Input
value = tf.constant([1, 15, 10], dtype = tf.float64)
  
# Printing the Input
print("Value: ", value)
  
# Converting Tensor to TensorProto
proto = tf.make_tensor_proto(value)
  
# Generating numpy array
res = tf.make_ndarray(proto)
  
# Printing the resulting numpy array
print("Result: ", res)

输出：

Value:  tf.Tensor([ 1. 15. 10.], shape=(3, ), dtype=float64)
Result:  [ 1. 15. 10.]

示例 2：此示例使用形状为 (2, 2) 的张量，因此结果数组的形状将为 (2, 2)。

Python3

# importing the library
import tensorflow as tf
  
# Initializing Input
value = tf.constant([[1, 2], [3, 4]], dtype = tf.float64)
  
# Printing the Input
print("Value: ", value)
  
# Converting Tensor to TensorProto
proto = tf.make_tensor_proto(value)
  
# Generating numpy array
res = tf.make_ndarray(proto)
  
# Printing the resulting numpy array
print("Result: ", res)

输出：

Value:  tf.Tensor(
[[1. 2.]
 [3. 4.]], shape=(2, 2), dtype=float64)
Result:  [[1. 2.]
 [3. 4.]]