How do I Convert TensorFlow Tensors to NumPy Arrays?

Converting Tensor to NumPy Array in TensorFlow

TensorFlow provides a convenient method, .numpy(), to convert tensors into NumPy arrays.

TensorFlow 2.x

Enabling eager execution makes TensorFlow operations immediately executable, allowing you to call .numpy() directly on tensors:

<code class="python">import tensorflow as tf

a = tf.constant([[1, 2], [3, 4]])
b = tf.add(a, 1)

print(a.numpy())
# array([[1, 2],
#        [3, 4]], dtype=int32)

print(b.numpy())
# array([[2, 3],
#        [4, 5]], dtype=int32)</code>

TensorFlow 1.x

If eager execution is disabled in TensorFlow 1.x, you can create a graph and execute it to obtain the NumPy array:

<code class="python">a = tf.constant([[1, 2], [3, 4]])
b = tf.add(a, 1)
out = tf.multiply(a, b)

with tf.compat.v1.Session() as sess:
    print(sess.run(out))
# array([[ 2,  6],
#        [12, 20]], dtype=int32)</code>

Notes:

• .numpy() may share memory with the original tensor, so changes to one may affect the other.
• If you encounter the error "AttributeError: 'Tensor' object has no attribute 'numpy'," ensure TF 2.0 is properly installed, or enable eager execution.

The above is the detailed content of How do I Convert TensorFlow Tensors to NumPy Arrays?. For more information, please follow other related articles on the PHP Chinese website!