解决Keras TensorFlow 混编中 trainable=False设置无效问题


Posted in Python onJune 28, 2020

这是最近碰到一个问题,先描述下问题:

首先我有一个训练好的模型(例如vgg16),我要对这个模型进行一些改变,例如添加一层全连接层,用于种种原因,我只能用TensorFlow来进行模型优化,tf的优化器,默认情况下对所有tf.trainable_variables()进行权值更新,问题就出在这,明明将vgg16的模型设置为trainable=False,但是tf的优化器仍然对vgg16做权值更新

以上就是问题描述,经过谷歌百度等等,终于找到了解决办法,下面我们一点一点的来复原整个问题。

trainable=False 无效

首先,我们导入训练好的模型vgg16,对其设置成trainable=False

from keras.applications import VGG16
import tensorflow as tf
from keras import layers
# 导入模型
base_mode = VGG16(include_top=False)
# 查看可训练的变量
tf.trainable_variables()
[<tf.Variable 'block1_conv1/kernel:0' shape=(3, 3, 3, 64) dtype=float32_ref>,
 <tf.Variable 'block1_conv1/bias:0' shape=(64,) dtype=float32_ref>,
 <tf.Variable 'block1_conv2/kernel:0' shape=(3, 3, 64, 64) dtype=float32_ref>,
 <tf.Variable 'block1_conv2/bias:0' shape=(64,) dtype=float32_ref>,
 <tf.Variable 'block2_conv1/kernel:0' shape=(3, 3, 64, 128) dtype=float32_ref>,
 <tf.Variable 'block2_conv1/bias:0' shape=(128,) dtype=float32_ref>,
 <tf.Variable 'block2_conv2/kernel:0' shape=(3, 3, 128, 128) dtype=float32_ref>,
 <tf.Variable 'block2_conv2/bias:0' shape=(128,) dtype=float32_ref>,
 <tf.Variable 'block3_conv1/kernel:0' shape=(3, 3, 128, 256) dtype=float32_ref>,
 <tf.Variable 'block3_conv1/bias:0' shape=(256,) dtype=float32_ref>,
 <tf.Variable 'block3_conv2/kernel:0' shape=(3, 3, 256, 256) dtype=float32_ref>,
 <tf.Variable 'block3_conv2/bias:0' shape=(256,) dtype=float32_ref>,
 <tf.Variable 'block3_conv3/kernel:0' shape=(3, 3, 256, 256) dtype=float32_ref>,
 <tf.Variable 'block3_conv3/bias:0' shape=(256,) dtype=float32_ref>,
 <tf.Variable 'block4_conv1/kernel:0' shape=(3, 3, 256, 512) dtype=float32_ref>,
 <tf.Variable 'block4_conv1/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block4_conv2/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block4_conv2/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block4_conv3/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block4_conv3/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block5_conv1/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block5_conv1/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block5_conv2/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block5_conv2/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block5_conv3/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block5_conv3/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block1_conv1_1/kernel:0' shape=(3, 3, 3, 64) dtype=float32_ref>,
 <tf.Variable 'block1_conv1_1/bias:0' shape=(64,) dtype=float32_ref>,
 <tf.Variable 'block1_conv2_1/kernel:0' shape=(3, 3, 64, 64) dtype=float32_ref>,
 <tf.Variable 'block1_conv2_1/bias:0' shape=(64,) dtype=float32_ref>,
 <tf.Variable 'block2_conv1_1/kernel:0' shape=(3, 3, 64, 128) dtype=float32_ref>,
 <tf.Variable 'block2_conv1_1/bias:0' shape=(128,) dtype=float32_ref>,
 <tf.Variable 'block2_conv2_1/kernel:0' shape=(3, 3, 128, 128) dtype=float32_ref>,
 <tf.Variable 'block2_conv2_1/bias:0' shape=(128,) dtype=float32_ref>,
 <tf.Variable 'block3_conv1_1/kernel:0' shape=(3, 3, 128, 256) dtype=float32_ref>,
 <tf.Variable 'block3_conv1_1/bias:0' shape=(256,) dtype=float32_ref>,
 <tf.Variable 'block3_conv2_1/kernel:0' shape=(3, 3, 256, 256) dtype=float32_ref>,
 <tf.Variable 'block3_conv2_1/bias:0' shape=(256,) dtype=float32_ref>,
 <tf.Variable 'block3_conv3_1/kernel:0' shape=(3, 3, 256, 256) dtype=float32_ref>,
 <tf.Variable 'block3_conv3_1/bias:0' shape=(256,) dtype=float32_ref>,
 <tf.Variable 'block4_conv1_1/kernel:0' shape=(3, 3, 256, 512) dtype=float32_ref>,
 <tf.Variable 'block4_conv1_1/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block4_conv2_1/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block4_conv2_1/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block4_conv3_1/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block4_conv3_1/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block5_conv1_1/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block5_conv1_1/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block5_conv2_1/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block5_conv2_1/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block5_conv3_1/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block5_conv3_1/bias:0' shape=(512,) dtype=float32_ref>]
# 设置 trainable=False
# base_mode.trainable = False似乎也是可以的
for layer in base_mode.layers:
  layer.trainable = False

设置好trainable=False后,再次查看可训练的变量,发现并没有变化,也就是说设置无效

# 再次查看可训练的变量
tf.trainable_variables()

[<tf.Variable 'block1_conv1/kernel:0' shape=(3, 3, 3, 64) dtype=float32_ref>,
 <tf.Variable 'block1_conv1/bias:0' shape=(64,) dtype=float32_ref>,
 <tf.Variable 'block1_conv2/kernel:0' shape=(3, 3, 64, 64) dtype=float32_ref>,
 <tf.Variable 'block1_conv2/bias:0' shape=(64,) dtype=float32_ref>,
 <tf.Variable 'block2_conv1/kernel:0' shape=(3, 3, 64, 128) dtype=float32_ref>,
 <tf.Variable 'block2_conv1/bias:0' shape=(128,) dtype=float32_ref>,
 <tf.Variable 'block2_conv2/kernel:0' shape=(3, 3, 128, 128) dtype=float32_ref>,
 <tf.Variable 'block2_conv2/bias:0' shape=(128,) dtype=float32_ref>,
 <tf.Variable 'block3_conv1/kernel:0' shape=(3, 3, 128, 256) dtype=float32_ref>,
 <tf.Variable 'block3_conv1/bias:0' shape=(256,) dtype=float32_ref>,
 <tf.Variable 'block3_conv2/kernel:0' shape=(3, 3, 256, 256) dtype=float32_ref>,
 <tf.Variable 'block3_conv2/bias:0' shape=(256,) dtype=float32_ref>,
 <tf.Variable 'block3_conv3/kernel:0' shape=(3, 3, 256, 256) dtype=float32_ref>,
 <tf.Variable 'block3_conv3/bias:0' shape=(256,) dtype=float32_ref>,
 <tf.Variable 'block4_conv1/kernel:0' shape=(3, 3, 256, 512) dtype=float32_ref>,
 <tf.Variable 'block4_conv1/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block4_conv2/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block4_conv2/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block4_conv3/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block4_conv3/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block5_conv1/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block5_conv1/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block5_conv2/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block5_conv2/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block5_conv3/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block5_conv3/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block1_conv1_1/kernel:0' shape=(3, 3, 3, 64) dtype=float32_ref>,
 <tf.Variable 'block1_conv1_1/bias:0' shape=(64,) dtype=float32_ref>,
 <tf.Variable 'block1_conv2_1/kernel:0' shape=(3, 3, 64, 64) dtype=float32_ref>,
 <tf.Variable 'block1_conv2_1/bias:0' shape=(64,) dtype=float32_ref>,
 <tf.Variable 'block2_conv1_1/kernel:0' shape=(3, 3, 64, 128) dtype=float32_ref>,
 <tf.Variable 'block2_conv1_1/bias:0' shape=(128,) dtype=float32_ref>,
 <tf.Variable 'block2_conv2_1/kernel:0' shape=(3, 3, 128, 128) dtype=float32_ref>,
 <tf.Variable 'block2_conv2_1/bias:0' shape=(128,) dtype=float32_ref>,
 <tf.Variable 'block3_conv1_1/kernel:0' shape=(3, 3, 128, 256) dtype=float32_ref>,
 <tf.Variable 'block3_conv1_1/bias:0' shape=(256,) dtype=float32_ref>,
 <tf.Variable 'block3_conv2_1/kernel:0' shape=(3, 3, 256, 256) dtype=float32_ref>,
 <tf.Variable 'block3_conv2_1/bias:0' shape=(256,) dtype=float32_ref>,
 <tf.Variable 'block3_conv3_1/kernel:0' shape=(3, 3, 256, 256) dtype=float32_ref>,
 <tf.Variable 'block3_conv3_1/bias:0' shape=(256,) dtype=float32_ref>,
 <tf.Variable 'block4_conv1_1/kernel:0' shape=(3, 3, 256, 512) dtype=float32_ref>,
 <tf.Variable 'block4_conv1_1/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block4_conv2_1/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block4_conv2_1/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block4_conv3_1/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block4_conv3_1/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block5_conv1_1/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block5_conv1_1/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block5_conv2_1/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block5_conv2_1/bias:0' shape=(512,) dtype=float32_ref>,
 <tf.Variable 'block5_conv3_1/kernel:0' shape=(3, 3, 512, 512) dtype=float32_ref>,
 <tf.Variable 'block5_conv3_1/bias:0' shape=(512,) dtype=float32_ref>]

解决的办法

解决的办法就是在导入模型的时候建立一个variable_scope,将需要训练的变量放在另一个variable_scope,然后通过tf.get_collection获取需要训练的变量,最后通过tf的优化器中var_list指定需要训练的变量

from keras import models
with tf.variable_scope('base_model'):
  base_model = VGG16(include_top=False, input_shape=(224,224,3))
with tf.variable_scope('xxx'):
  model = models.Sequential()
  model.add(base_model)
  model.add(layers.Flatten())
  model.add(layers.Dense(10))
# 获取需要训练的变量
trainable_var = tf.get_collection(tf.GraphKeys.TRAINABLE_VARIABLES, 'xxx')
trainable_var

[<tf.Variable 'xxx_2/dense_1/kernel:0' shape=(25088, 10) dtype=float32_ref>,
<tf.Variable 'xxx_2/dense_1/bias:0' shape=(10,) dtype=float32_ref>]

# 定义tf优化器进行训练,这里假设有一个loss
loss = model.output / 2; # 随便定义的,方便演示
train_step = tf.train.AdamOptimizer().minimize(loss, var_list=trainable_var)

总结

在keras与TensorFlow混编中,keras中设置trainable=False对于TensorFlow而言并不起作用

解决的办法就是通过variable_scope对变量进行区分,在通过tf.get_collection来获取需要训练的变量,最后通过tf优化器中var_list指定训练

以上这篇解决Keras TensorFlow 混编中 trainable=False设置无效问题就是小编分享给大家的全部内容了,希望能给大家一个参考,也希望大家多多支持三水点靠木。

Python 相关文章推荐
Windows下python2.7.8安装图文教程
May 26 Python
详谈python在windows中的文件路径问题
Apr 28 Python
python中计算一个列表中连续相同的元素个数方法
Jun 29 Python
Python面向对象之继承和组合用法实例分析
Aug 27 Python
Python中修改字符串的四种方法
Nov 02 Python
使用Django开发简单接口实现文章增删改查
May 09 Python
python 实现兔子生兔子示例
Nov 21 Python
Django密码存储策略分析
Jan 09 Python
Python continue语句实例用法
Feb 06 Python
Python3 字典dictionary入门基础附实例
Feb 10 Python
PyCharm2020.1.2社区版安装,配置及使用教程详解(Windows)
Aug 07 Python
Python学习之time模块的基本使用
Jan 17 Python
Keras: model实现固定部分layer,训练部分layer操作
Jun 28 #Python
sklearn的predict_proba使用说明
Jun 28 #Python
基于python实现ROC曲线绘制广场解析
Jun 28 #Python
Python sklearn中的.fit与.predict的用法说明
Jun 28 #Python
浅谈sklearn中predict与predict_proba区别
Jun 28 #Python
解决Pytorch自定义层出现多Variable共享内存错误问题
Jun 28 #Python
Pytorch学习之torch用法----比较操作(Comparison Ops)
Jun 28 #Python
You might like
php常用的url处理函数总结
2014/11/19 PHP
php+mysqli事务控制实现银行转账实例
2015/01/29 PHP
jQuery向下滚动即时加载内容实现的瀑布流效果
2016/01/07 PHP
微信公众号支付之坑:调用支付jsapi缺少参数 timeStamp等错误解决方法
2016/01/12 PHP
javascript eval和JSON之间的联系
2009/12/31 Javascript
纯js实现无限空间大小的本地存储
2015/06/18 Javascript
jQuery实现按钮点击遮罩加载及处理完后恢复的效果
2016/06/07 Javascript
Bootstrap树形菜单插件TreeView.js使用方法详解
2016/11/01 Javascript
js实现随机点名系统(实例讲解)
2017/10/18 Javascript
微信小程序实现下拉刷新和轮播图效果
2017/11/21 Javascript
vue底部加载更多的实例代码
2018/06/29 Javascript
Bootstrap的aria-label和aria-labelledby属性实例详解
2018/11/02 Javascript
详解js实时获取并显示当前时间的方法
2019/05/10 Javascript
解决微信浏览器缓存站点入口文件(IIS部署Vue项目)
2019/06/17 Javascript
Vue.directive 实现元素scroll逻辑复用
2019/11/29 Javascript
在vue-cli3中使用axios获取本地json操作
2020/07/30 Javascript
Vue优化:常见会导致内存泄漏问题及优化详解
2020/08/04 Javascript
JS中箭头函数与this的写法和理解
2021/01/14 Javascript
[05:05]第三天的dota2
2013/07/29 DOTA
[04:05]TI9战队采访 - Natus Vincere
2019/08/22 DOTA
windows下python和pip安装教程
2018/05/25 Python
强悍的Python读取大文件的解决方案
2019/02/16 Python
python3 selenium自动化 下拉框定位的例子
2019/08/23 Python
PyQt5中多线程模块QThread使用方法的实现
2020/01/31 Python
如何利用python读取micaps文件详解
2020/10/18 Python
Vivo俄罗斯官方在线商店:中国智能手机品牌
2019/10/04 全球购物
建筑人员岗位职责
2013/12/25 职场文书
结婚典礼证婚词
2014/01/11 职场文书
党支部书记岗位责任制
2014/02/11 职场文书
物业管理专业求职信
2014/06/11 职场文书
羽毛球比赛策划方案
2014/06/13 职场文书
2014县政府领导班子三严三实对照检查材料思想汇报
2014/09/26 职场文书
部门优秀员工推荐信
2015/03/24 职场文书
原料仓管员岗位职责
2015/04/01 职场文书
Python实现的扫码工具居然这么好用!
2021/06/07 Python
修改并编译golang源码的操作步骤
2021/07/25 Golang