<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[veryverycold的博客]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/veryverycold</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; veryverycold]]></copyright><item><title><![CDATA[tf2.0keras一个简单的自定义层]]></title><link>https://blog.youkuaiyun.com/veryverycold/article/details/106524274</link><guid>https://blog.youkuaiyun.com/veryverycold/article/details/106524274</guid><author>veryverycold</author><pubDate>Wed, 03 Jun 2020 14:56:44 +0800</pubDate><description><![CDATA[自定义层





init()
首先是一些必要的参数的初始化


build()
主要声明需要更新的参数部分


call()
包含了主要层的实现


class MyLayer(tf.keras.layers.Layer):
    def __init__(self,output_dim):
    	#接收传的值
        self.output_dim=output_dim
        super(MyLayer,self).__init__()
        
    def...]]></description><category></category></item><item><title><![CDATA[一个简单的keras的模型创建,编译与优化器]]></title><link>https://blog.youkuaiyun.com/veryverycold/article/details/106493440</link><guid>https://blog.youkuaiyun.com/veryverycold/article/details/106493440</guid><author>veryverycold</author><pubDate>Tue, 02 Jun 2020 12:03:57 +0800</pubDate><description><![CDATA[import tensorflow as tf
import tensorflow.keras.layers as layers
import tensorflow.keras.optimizer as optimizer
#神经网络的模型构建
model=tf.keras.Sequential()
model.add(layers.Dense(256,activation="relu"))
model.add(layers.Dense(128,activation="relu"))
model.add(l]]></description><category></category></item><item><title><![CDATA[tensorflow低阶api基础编程]]></title><link>https://blog.youkuaiyun.com/veryverycold/article/details/106488762</link><guid>https://blog.youkuaiyun.com/veryverycold/article/details/106488762</guid><author>veryverycold</author><pubDate>Tue, 02 Jun 2020 11:14:23 +0800</pubDate><description><![CDATA[tf常量
import tensorflow as tf
import numpy as np
#tf常量的定义
a=tf.constant(7)
#tf常量的值
print(a.numpy())
#python中tf常量输出
print(a)


#结果
7
tf.Tensor(7, shape=(), dtype=int32)

tf变量
#声明一个tf变量
a2=tf.Variable(7)
#声明一个tf变量数组
a3=tf.Variable([1,2,3])
print(a2,a3)
print(]]></description><category></category></item><item><title><![CDATA[为了效率!!!TensorFLow2.0静态图与eager模式(动态图)]]></title><link>https://blog.youkuaiyun.com/veryverycold/article/details/106471154</link><guid>https://blog.youkuaiyun.com/veryverycold/article/details/106471154</guid><author>veryverycold</author><pubDate>Mon, 01 Jun 2020 15:42:51 +0800</pubDate><description><![CDATA[总结自此博客
为了效率!!!eager与静态图转换的那些坑




eager模式:
就是动态图




调用方法:
tensorflow2.0默认eager模式


优点
Python写法,方便调试


缺点
比自动图速度慢






静  态 图 模 式 :





调用方法:
需要在函数上方加上@tf.function装饰


优点
tf写法,速度快


缺点
不如python写法直观






转换方式
在头上加@tf.function



第一个大坑
#调用静态图的例子
@tf.funct]]></description><category></category></item><item><title><![CDATA[H5Css  float:left居中]]></title><link>https://blog.youkuaiyun.com/veryverycold/article/details/106408311</link><guid>https://blog.youkuaiyun.com/veryverycold/article/details/106408311</guid><author>veryverycold</author><pubDate>Thu, 28 May 2020 16:42:37 +0800</pubDate><description><![CDATA[如何居中float:left的元素
pading,margin,border全设置为0,经过嵌套div可以实现
如果想设置多float:left水平居中的效果,可以设置子div的宽度百分比,与margin百分比实现




]]></description><category></category></item></channel></rss>