文章目录
一、简介
slim
被放在tensorflow.contrib
这个库下面,引入的方法如下:
import tensorflow.contrib.slim as slim
或
import tensorflow as tf
slim = tf.contrib.slim
那么什么是 slim ?slim到底有什么用 ?
它可以消除原生tensorflow
里面很多重复的模板性的代码,让代码更精确,更容易替换性。另外slim
提供了很多计算机视觉方面的著名模型(VGG,AlexNet等),我们不仅可以直接使用,甚至能以各种方式进行扩展。
二、slim的子模块及功能介绍
arg_scope:提供了一个名为arg_scope的新范围,该范围允许用户为该范围内的特定操作定义默认参数。
除了基本的 namescope,variabelscope外,又加了argscope,它是用来控制每一层的替代超参数的。
data(数据):包含TF-slim的数据集定义,数据提供程序,parallel_reader和解码实用程序。
evaluation(评估):包含评估模型的例程,评估模型的一些方法,用的不多
layers(层):包含用于使用张量流构建模型的高层。
这个比较重要,slim的核心和精髓,一些复杂层的定义
learning(学习):包含用于训练模型的模型
loss:包含常用的损失函数。
matris(指标):包含流行的评估指标,评估模型的准则标准
nets(网络):包含流行的网络定义,例如VGG和AlexNet模型。
queues(队列):提供上下文管理器,可轻松安全地启动和关闭QueueRunner。
variables(变量):为变量创建和操作提供方便的包装。
三、slim 定义模型
1. slim中定义一个变量的示例
# Model Variables
weights = slim.model_variable('weights',
shape=[10, 10, 3 , 3],
initializer=tf.truncated_normal_initializer(stddev=0.1),
regularizer=slim.l2_regularizer(0.05),
device='/CPU:0')
model_variables = slim.get_model_variables()
# Regular variables
my_var = slim.variable('my_var',
shape=[20, 1],
initializer=tf.zeros_initializer())
regular_variables_and_model_variables = slim.get_variables()
如上,变量分为两类:模型变量和局部变量。
局部变量是不作为模型参数保存的,而模型变量会在保存的时候保存下来。这个玩过tensorflow的人都会明白,诸如global_step之类的就是局部变量。
slim中可以写明变量存放的设备,正则和初始化规则。
还有获取变量的函数也需要注意一下,get_variables是返回所有的变量。
2. slim中实现一个层
首先让我们看看tensorflow怎么实现一个层,例如卷积层:
input = ...
with tf.name_scope('conv1_1') as scope:
kernel = tf.Variable(tf.truncated_normal([3, 3, 64, 128], dtype=tf.float32,
stddev=1e-1), name='weights')
conv = tf.nn.conv2d(input, kernel, [1, 1, 1, 1], padding='SAME')
biases = tf.Variable(tf.constant(0.0, shape=[128], dtype=tf.float32),
trainable=True, name='biases')
bias = tf.nn.bias_add(conv, biases)
conv1 = tf.nn.relu(bias, name=scope)
然后slim的实现:
input = ...
net = slim.conv2d(input, 128, [3, 3], scope='conv1_1')
但这个不是重要的,因为 tenorflow 目前也有大部分层的简单实现,这里比较吸引人的是 slim 中的 repeat 和 stack 操作:
假设定义三个相同的卷积层:
net = ...
net = slim.conv2d(net, 256, [3, 3], scope='conv3_1')
net = slim.conv2d(net, 256, [3, 3], scope='conv3_2')
net = slim.conv2d(net, 256, [3, 3], scope='conv3_3')
net = slim.max_pool2d(net, [2, 2], scope='pool2')
在slim中的repeat操作可以减少代码量:
net = slim.repeat(net, 3, slim.conv2d, 256, [3, 3], scope='conv3')
net = slim.max_pool2d(net, [2, 2], scope='pool2')
假设定义三层FC:
# Verbose way:
x = slim.fully_connected(x, 32, scope='fc/fc_1')
x = slim.fully_connected(x, 64, scope='fc/fc_2')
x = slim.fully_connected(x, 128, scope='fc/fc_3')
使用stack操作:
slim.stack(x, slim.fully_connected, [32, 64, 128], scope='fc')
同理卷积层也一样:
# 普通方法:
x = slim.conv2d(x, 32, [3, 3], scope='core/core_1')
x = slim.conv2d(x, 32, [1, 1], scope='core/core_2')
x = slim.conv2d(x, 64, [3, 3], scope='core/core_3')
x = slim.conv2d(x, 64, [1, 1], scope='core/core_4')
# 简便方法:
slim.stack(x, slim.conv2d, [(32, [3, 3]), (32, [1, 1]), (64, [3, 3]), (64, [1, 1])], scope='core')
3. slim 中的 argscope
如果你的网络有大量相同的参数,如下:
net = slim.conv2d(inputs, 64, [11, 11], 4, padding='SAME',
weights_initializer=tf.truncated_normal_initializer(stddev=0.01),
weights_regularizer=slim.l2_regularizer(0.0005), scope='conv1')
net = slim.conv2d(net, 128, [11, 11], padding='VALID',
weights_initializer=tf.truncated_normal_initializer(stddev=0.01),
weights_regularizer=slim.l2_regularizer(0.0005), scope='conv2')
net = slim.conv2d(net, 256, [11, 11], padding='SAME',
weights_initializer=tf.truncated_normal_initializer(stddev=0.01),
weights_regularizer=slim.l2_regularizer(0.0005), scope='conv3')
然后我们用arg_scope处理一下:
with slim.arg_scope([slim.conv2d], padding='SAME',
weights_initializer=tf.truncated_normal_initializer(stddev=0.01)
weights_regularizer=slim.l2_regularizer(0.0005)):
net = slim.conv2d(inputs, 64, [11, 11], scope='conv1')
net = slim.conv2d(net, 128, [11, 11], padding='VALID', scope='conv2')
net = slim.conv2d(net, 256, [11, 11], scope='conv3')
是不是一下子就变简洁了?
这里额外说明一点,arg_scope 的作用范围内,是定义了指定层的替代参数,若想特别指定某些层的参数,可以重新赋值(相当于重置),如上倒数第二行代码。
那如果除了卷积层还有其他层呢?那就要如下定义:
with slim.arg_scope([slim.conv2d, slim.fully_connected],
activation_fn=tf.nn.relu,
weights_initializer=tf.truncated_normal_initializer(stddev=0.01),
weights_regularizer=slim.l2_regularizer(0.0005)):
with slim.arg_scope([slim.conv2d], stride=1, padding='SAME'):
net = slim.conv2d(inputs, 64, [11, 11], 4, padding='VALID', scope='conv1')
net = slim.conv2d(net, 256, [5, 5],
weights_initializer=tf.truncated_normal_initializer(stddev=0.03),
scope='conv2')
net = slim.fully_connected(net, 1000, activation_fn=None, scope='fc')
写两个arg_scope就行了
采用如上方法,定义一个VGG也就十几行代码的事了
def vgg16(inputs):
with slim.arg_scope([slim.conv2d, slim.fully_connected],
activation_fn=tf.nn.relu,
weights_initializer=tf.truncated_normal_initializer(0.0, 0.01),
weights_regularizer=slim.l2_regularizer(0.0005)):
net = slim.repeat(inputs, 2, slim.conv2d, 64, [3, 3], scope='conv1')
net = slim.max_pool2d(net, [2, 2], scope='pool1')
net = slim.repeat(net, 2, slim.conv2d, 128, [3, 3], scope='conv2')
net = slim.max_pool2d(net, [2, 2], scope='pool2')
net = slim.repeat(net, 3, slim.conv2d, 256, [3, 3], scope='conv3')
net = slim.max_pool2d(net, [2, 2], scope='pool3')
net = slim.repeat(net, 3, slim.conv2d, 512, [3, 3], scope='conv4')
net = slim.max_pool2d(net, [2, 2], scope='pool4')
net = slim.repeat(net, 3, slim.conv2d, 512, [3, 3], scope='conv5')
net = slim.max_pool2d(net, [2, 2], scope='pool5')
net = slim.fully_connected(net, 4096, scope='fc6')
net = slim.dropout(net, 0.5, scope='dropout6')
net = slim.fully_connected(net, 4096, scope='fc7')
net = slim.dropout(net, 0.5, scope='dropout7')
net = slim.fully_connected(net, 1000, activation_fn=None, scope='fc8')
return net
四、训练模型
这个没什么好说的,说一下直接拿经典网络来训练吧。
import tensorflow as tf
vgg = tf.contrib.slim.nets.vgg
# Load the images and labels.
images, labels = ...
# Create the model.
predictions, _ = vgg.vgg_16(images)
# Define the loss functions and get the total loss.
loss = slim.losses.softmax_cross_entropy(predictions, labels)
不是超级简单?
关于损耗,要说一下定义自己的损耗的方法,以及注意不要忘记加入到slim中让slim看到你的损耗。
还有正则项也是需要手动添加进损当中的,不然最后计算的时候就不优化正则目标了。
# Load the images and labels.
images, scene_labels, depth_labels, pose_labels = ...
# Create the model.
scene_predictions, depth_predictions, pose_predictions = CreateMultiTaskModel(images)
# Define the loss functions and get the total loss.
classification_loss = slim.losses.softmax_cross_entropy(scene_predictions, scene_labels)
sum_of_squares_loss = slim.losses.sum_of_squares(depth_predictions, depth_labels)
pose_loss = MyCustomLossFunction(pose_predictions, pose_labels)
slim.losses.add_loss(pose_loss) # Letting TF-Slim know about the additional loss.
# The following two ways to compute the total loss are equivalent:
regularization_loss = tf.add_n(slim.losses.get_regularization_losses())
total_loss1 = classification_loss + sum_of_squares_loss + pose_loss + regularization_loss
# (Regularization Loss is included in the total loss by default).
total_loss2 = slim.losses.get_total_loss()
五、读取保存模型变量
通过以下功能我们可以加载模型的部分变量:
# Create some variables.
v1 = slim.variable(name="v1", ...)
v2 = slim.variable(name="nested/v2", ...)
...
# Get list of variables to restore (which contains only 'v2').
variables_to_restore = slim.get_variables_by_name("v2")
# Create the saver which will be used to restore the variables.
restorer = tf.train.Saver(variables_to_restore)
with tf.Session() as sess:
# Restore variables from disk.
restorer.restore(sess, "/tmp/model.ckpt")
print("Model restored.")
除了这种部分变量加载的方法外,我们甚至还能加载到不同的名字的变量中。
假设我们定义的网络变量是conv1 / weights
,而从VGG加载的变量称为vgg16 / conv1 / weights
,正常负载肯定会报错(找不到变量名),但是可以这样:
def name_in_checkpoint(var):
return 'vgg16/' + var.op.name
variables_to_restore = slim.get_model_variables()
variables_to_restore = {name_in_checkpoint(var):var for var in variables_to_restore}
restorer = tf.train.Saver(variables_to_restore)
with tf.Session() as sess:
# Restore variables from disk.
restorer.restore(sess, "/tmp/model.ckpt")
通过这种方式我们可以加载不同变量名的变量!!
参考:
【1】https://blog.youkuaiyun.com/zhoujunr1/article/details/77131605
【2】https://www.2cto.com/kf/201706/649266.html
【3】https://blog.youkuaiyun.com/zsWang9/article/details/79965501