Python – tensorflow.GradientTape.watch()
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。
watch()用于通过 Tape 开始跟踪张量。
Syntax: watch( tensor )
Parameter:
- tensor: It is a Tensor or list of tensors to be watched.
Returns: None
Raise:
- ValueError: It will raise ValueError if the passes parameter is not Tensor.
示例 1:
Python3
# Importing the library
import tensorflow as tf
x = tf.constant(4.0)
# Using GradientTape
with tf.GradientTape() as gfg:
# Starting the recording x
gfg.watch(x)
y = x * x
# Computing gradient
res = gfg.gradient(y, x)
# Printing result
print("res: ", res)
Python3
# Importing the library
import tensorflow as tf
x = tf.constant(4.0)
z = tf.constant(5.0)
# Using GradientTape
with tf.GradientTape(persistent = True) as gfg:
# Starting the recording x and z
gfg.watch([x, z])
y = z * z
u = x * x
# Computing gradient
grad_y = gfg.gradient(y, z)
grad_u = gfg.gradient(u, x)
# Printing result
print("grad_y: ", grad_y)
print("grad_u: ", grad_u)
输出:
res: tf.Tensor(8.0, shape=(), dtype=float32)
示例 2:
Python3
# Importing the library
import tensorflow as tf
x = tf.constant(4.0)
z = tf.constant(5.0)
# Using GradientTape
with tf.GradientTape(persistent = True) as gfg:
# Starting the recording x and z
gfg.watch([x, z])
y = z * z
u = x * x
# Computing gradient
grad_y = gfg.gradient(y, z)
grad_u = gfg.gradient(u, x)
# Printing result
print("grad_y: ", grad_y)
print("grad_u: ", grad_u)
输出:
grad_y: tf.Tensor(10.0, shape=(), dtype=float32)
grad_u: tf.Tensor(8.0, shape=(), dtype=float32)