自己定制训练

1，以类的方式定义一个模型

class Model(object):
def __init__(self):
# Initialize variable to (5.0, 0.0)
# In practice, these should be initialized to random values.
self.W = tf.Variable(5.0)
self.b = tf.Variable(0.0)

def __call__(self, x):
return self.W * x + self.b

model = Model()

assert model(3.0).numpy() == 15.0

2，损失函数

def loss(predicted_y, desired_y):
return tf.reduce_mean(tf.square(predicted_y - desired_y))

3，生成数据

TRUE_W = 3.0
TRUE_b = 2.0
NUM_EXAMPLES = 1000

inputs = tf.random_normal(shape=[NUM_EXAMPLES])
noise = tf.random_normal(shape=[NUM_EXAMPLES])
outputs = inputs * TRUE_W + TRUE_b + noise

4，绘制，训练前

import matplotlib.pyplot as plt

plt.scatter(inputs, outputs, c='b',s=1)
plt.scatter(inputs, model(inputs), c='r',linewidths=0.01)
plt.show()

print('Current loss: '),
print(loss(model(inputs), outputs).numpy())

5，迭代过程

def train(model, inputs, outputs, learning_rate):
　　with tf.GradientTape() as t:
　　　　current_loss = loss(model(inputs), outputs)
　　dW, db = t.gradient(current_loss, [model.W, model.b])
　　model.W.assign_sub(learning_rate * dW)
　　model.b.assign_sub(learning_rate * db)

6，训练过程

model = Model()

# Collect the history of W-values and b-values to plot later
Ws, bs = [], []
epochs = range(10)
for epoch in epochs:
Ws.append(model.W.numpy())
bs.append(model.b.numpy())
current_loss = loss(model(inputs), outputs)

train(model, inputs, outputs, learning_rate=0.1)
print('Epoch %2d: W=%1.2f b=%1.2f, loss=%2.5f' %
(epoch, Ws[-1], bs[-1], current_loss))

# Let's plot it all
plt.plot(epochs, Ws, 'r',
epochs, bs, 'b')
plt.plot([TRUE_W] * len(epochs), 'r--',
[TRUE_b] * len(epochs), 'b--')
plt.legend(['W', 'b', 'true W', 'true_b'])
plt.show()