1、tf.app.flags
tf定义了tf.app.flags,用于支持接受命令行传递参数,相当于接受argv。tf.app.flags.DEFINE_xxx()就是添加命令行的optional argument(可选参数),而tf.app.flags.FLAGS可以从对应的命令行参数取出参数。
import tensorflow as tf # 第一个是参数名称,第二个参数是默认值,第三个是参数描述 tf.app.flags.DEFINE_float('float_name', 0.01, 'input a float') tf.app.flags.DEFINE_string('str_name', 'def_v_1', "descrip1") tf.app.flags.DEFINE_integer('int_name', 10, "descript2") tf.app.flags.DEFINE_boolean('bool_name', False, "descript3") FLAGS = tf.app.flags.FLAGS # 必须带参数,否则:'TypeError: main() takes no arguments (1 given)'; main的参数名随意定义,无要求 def main(_): print(FLAGS.float_name) print(FLAGS.str_name) print(FLAGS.int_name) print(FLAGS.bool_name) if __name__ == '__main__': tf.app.run() # 执行main函数
执行:
(tf_learn) [@l_106 ~/ssd-balancap]$ python exc2.py 0.01 def_v_1 10 False (tf_learn) [@l_106 ~/ssd-balancap]$ python exc2.py --float_name 0.6 --str_name test_str --int_name 99 --bool_name True 0.6 test_str 99 True
2、slim
导入:
import tensorflow.contrib.slim as slim
arg_scope:用来控制每一层的默认超参数的。
定义变量
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,而模型变量会再save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。
定义卷积层:
input = [1,224,224,3] #tensorflow 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) #slim net = slim.conv2d(input, 128, [3, 3], scope='conv1_1')
repeat操作:
repeat操作可以减少代码量。
net = '' #原版 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') #repeat简化版 net = slim.repeat(net, 3, slim.conv2d, 256, [3, 3], scope='conv3') net = slim.max_pool2d(net, [2, 2], scope='pool2')
stack操作:
stack是处理卷积核或者输出不一样的情况。
#普通版 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') #stack简化版 slim.stack(x, slim.fully_connected, [32, 64, 128], scope='fc') #普通版: 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') #stack简化版: slim.stack(x, slim.conv2d, [(32, [3, 3]), (32, [1, 1]), (64, [3, 3]), (64, [1, 1])], scope='core')
argscope:
#普通版 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') #arg_scope简化版 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')
arg_scope的作用范围内,是定义了指定层的默认参数,若想特别指定某些层的参数,可以重新赋值(相当于重写),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: