tensorflow常用函数介绍
tensorflow tf.train.Supervisor作用
tf.train.Supervisor可以简化编程,避免显示地实现restore操作.通过一个例子看.
import tensorflow as tf
import numpy as np
import os
log_path = r"D:Sourcemodellinear"
log_name = "linear.ckpt"
# Create 100 phony x, y data points in NumPy, y = x * 0.1 + 0.3
x_data = np.random.rand(100).astype(np.float32)
y_data = x_data * 0.1 + 0.3
# Try to find values for W and b that compute y_data = W * x_data + b
# (We know that W should be 0.1 and b 0.3, but TensorFlow will
# figure that out for us.)
W = tf.Variable(tf.random_uniform([1], -1.0, 1.0))
b = tf.Variable(tf.zeros([1]))
y = W * x_data + b
# Minimize the mean squared errors.
loss = tf.reduce_mean(tf.square(y - y_data))
optimizer = tf.train.GradientDescentOptimizer(0.5)
train = optimizer.minimize(loss)
# Before starting, initialize the variables. We will 'run' this first.
saver = tf.train.Saver()
init = tf.global_variables_initializer()
# Launch the graph.
sess = tf.Session()
sess.run(init)
if len(os.listdir(log_path)) != 0: # 已经有模型直接读取
saver.restore(sess, os.path.join(log_path, log_name))
for step in range(201):
sess.run(train)
if step % 20 == 0:
print(step, sess.run(W), sess.run(b))
saver.save(sess, os.path.join(log_path, log_name))
这段代码是对tensorflow官网上的demo做一个微小的改动.如果模型已经存在,就先读取模型接着训练.tf.train.Supervisor可以简化这个步骤.看下面的代码.
import tensorflow as tf
import numpy as np
import os
log_path = r"D:Sourcemodelsupervisor"
log_name = "linear.ckpt"
x_data = np.random.rand(100).astype(np.float32)
y_data = x_data * 0.1 + 0.3
W = tf.Variable(tf.random_uniform([1], -1.0, 1.0))
b = tf.Variable(tf.zeros([1]))
y = W * x_data + b
loss = tf.reduce_mean(tf.square(y - y_data))
optimizer = tf.train.GradientDescentOptimizer(0.5)
train = optimizer.minimize(loss)
saver = tf.train.Saver()
init = tf.global_variables_initializer()
sv = tf.train.Supervisor(logdir=log_path, init_op=init) # logdir用来保存checkpoint和summary
saver = sv.saver # 创建saver
with sv.managed_session() as sess: # 会自动去logdir中去找checkpoint，如果没有的话，自动执行初始化
for i in range(201):
sess.run(train)
if i % 20 == 0:
print(i, sess.run(W), sess.run(b))
saver.save(sess, os.path.join(log_path, log_name))
sv = tf.train.Supervisor(logdir=log_path, init_op=init)会判断模型是否存在.如果存在,会自动读取模型.不用显式地调用restore.
tf.global_variables_initializer()
1.问题描述
很多时候，我们都会在会话中加入一句：
sess.run(tf.global_variables_initializer())
至于为什么，好像就是觉得要初始化变量而已，其实，这句话有具体的应用原因
2.用程序来说明
import tensorflow as tf
# 必须要使用global_variables_initializer的场合
# 含有tf.Variable的环境下，因为tf中建立的变量是没有初始化的，也就是在debug时还不是一个tensor量，而是一个Variable变量类型
size_out = 10
tensor = tf.Variable(tf.random_normal(shape=[size_out]))
init = tf.global_variables_initializer()
with tf.Session() as sess:
sess.run(init) # initialization variables
print(sess.run(tensor))
# 可以不适用初始化的场合
# 不含有tf.Variable、tf.get_Variable的环境下
# 比如只有tf.random_normal或tf.constant等
size_out = 10
tensor = tf.random_normal(shape=[size_out]) # 这里debug是一个tensor量哦
init = tf.global_variables_initializer()
with tf.Session() as sess:
# sess.run(init) # initialization variables
print(sess.run(tensor))
tf.gradients()
在tensorflow中，tf.gradients()的参数如下：
tf.gradients(ys, xs,
grad_ys=None,
name='gradients',
colocate_gradients_with_ops=False,
gate_gradients=False,
aggregation_method=None,
stop_gradients=None)
1234567
先不给出参数的意义~
对求导函数而言，其主要功能即求导公式：∂ y ∂ x frac {partial y}{partial x}∂x∂y。在tensorflow中，y yy和x xx都是tensor。
更进一步，tf.gradients()接受求导值ys和xs不仅可以是tensor，还可以是list，形如[tensor1, tensor2, …, tensorn]。当ys和xs都是list时，它们的求导关系为：
gradients()adds ops to the graph to output the derivatives ofyswith respect toxs. It returns a list ofTensorof lengthlen(xs)where each tensor is thesum(dy/dx)for y inys.
意思是：
-
tf.gradients()实现ys对xs求导 -
求导返回值是一个list，list的长度等于
len(xs) -
假设返回值是[grad1, grad2, grad3]，
ys=[y1, y2]，
xs=[x1, x2, x3]。则，真实的计算过程为:
- g r a d 1 = y 1 x 1 + y 2 x 1 grad1 = frac {y1}{x1} + frac {y2}{x1}gra**d1=x1y1+x1y2
- g r a d 2 = y 1 x 2 + y 2 x 2 grad2 = frac {y1}{x2} + frac {y2}{x2}gra**d2=x2y1+x2y2
- g r a d 3 = y 1 x 3 + y 2 x 3 grad3 = frac {y1}{x3} + frac {y2}{x3}gra**d3=x3y1+x3y2
基础实践
以线性回归为例，实践tf.gradients()的基础功能。线性回归：y = 3 × x + 2 y = 3 imes x + 2y=3×x+2
import numpy as np
import tensorflow as tf
sess = tf.Session()
x_input = tf.placeholder(tf.float32, name='x_input')
y_input = tf.placeholder(tf.float32, name='y_input')
w = tf.Variable(2.0, name='weight')
b = tf.Variable(1.0, name='biases')
y = tf.add(tf.multiply(x_input, w), b)
loss_op = tf.reduce_sum(tf.pow(y_input - y, 2)) / (2 * 32)
train_op = tf.train.GradientDescentOptimizer(0.01).minimize(loss_op)
gradients_node = tf.gradients(loss_op, w)
print(gradients_node)
'''tensorboard'''
# tf.summary.scalar('norm_grads', gradients_node)
# tf.summary.histogram('norm_grads', gradients_node)
# merged = tf.summary.merge_all()
# writer = tf.summary.FileWriter('log')
init = tf.global_variables_initializer()
sess.run(init)
'''构造数据集'''
x_pure = np.random.randint(-10, 100, 32)
x_train = x_pure + np.random.randn(32) / 10 # 为x加噪声
y_train = 3 * x_pure + 2 + np.random.randn(32) / 10 # 为y加噪声
for i in range(20):
_, gradients, loss = sess.run([train_op, gradients_node, loss_op],
feed_dict={x_input: x_train[i], y_input: y_train[i]})
print("epoch: {} loss: {} gradients: {}".format(i, loss, gradients))
sess.close()
123456789101112131415161718192021222324252627282930313233343536
输出：
epoch: 0 loss: 94.6083221436 gradients: [-187.66052]
epoch: 1 loss: 1.52120530605 gradients: [3.0984864]
epoch: 2 loss: 101.41834259 gradients: [241.91911]
...
epoch: 18 loss: 0.0215022582561 gradients: [-0.44370675]
epoch: 19 loss: 0.0189439821988 gradients: [-0.31349587]
123456
可以看到梯度逐渐减小，说明模型逐渐收敛。同时也可以看到参数更新的方向主要是梯度下降的方向（尽管伴随着震荡）。
tf.random_uniform
首先，TensorFlow官网给出的解释：
tf.random_normal(
shape,
mean=0.0,
stddev=1.0,
dtype=tf.float32,
seed=None,
name=None
)12345678
shape代表形状，也就是1纬的还是2纬的还是n纬的数组。
下面看图说话:
import tensorflow as tf
res = tf.random_uniform([1], -1, 1)
with tf.Session() as sess:
print(sess.run(res))
#结果为：[0.8457055]
123456
import tensorflow as tf
res = tf.random_uniform([2], -1, 1)
with tf.Session() as sess:
print(sess.run(res))
#结果为：[-0.20672345 0.6750064 ]12345
import tensorflow as tf
res = tf.random_uniform((4, 4), -1, 1)
with tf.Session() as sess:
print(sess.run(res))
#结果为：
[[ 0.61043835 -0.35983467 -0.02590227 0.6653023 ]
[ 0.4215083 0.6378925 -0.5907881 0.94997907]
[ 0.02292204 -0.7329526 0.03539038 -0.63158727]
[ 0.15353537 -0.21372676 -0.5452025 -0.44808888]]