Tag：link
-
Time：2020-5-25
preface When your R & D team is getting larger and larger, or you can’t stand the super slow construction of node, you can consider to read on and recommend a relatively complete solution based on verdaccio. Because of the environment, we go directlynpmjs.orgDon’t think about downloading. You cannpm config set registry=https://registry.npm.taobao.orgIt can alleviate some […]
-
Time：2020-5-25
The example of this article describes how PHP regular expressions can get all links of content. Share with you for your reference. The details are as follows: Here are two methods for PHP regular links. They can get all links of content and save them to an array. Of course, they can also replace all […]
-
Time：2020-5-25
This paper uses a simple Python crawler, mail sending and timing tasks to achieve dailySend bedtime stories regularlyIs a step-by-step article. After testing, the program can still run normally. Recently, some cute asked me to tell her a little story every night before going to bed after I finished my work. I thought that there […]
-
Time：2020-5-23
In this paper, based on the parameter record 10, zoom is added to the data enhancement part_ Range = 0.2, increase the number of training iterations to 5000 epoch, change the batch size to 625, and test the effect of adaptive parametric relu (aprelu) activation function on cifar10 image set. The principle of adaptive parametric […]
-
Time：2020-5-22
From the previous results of parameter adjustment, over fitting is the most important problem. In this paper, on the basis of parameter record 12, the number of layers is reduced to 9 residual modules, and try again. The principle of the adaptive parametric relu activation function is as follows: The keras program is as follows: […]
-
Time：2020-5-22
Download PostgreSQL at this link. When installing, remember the initial password you set for user Postgres and the default port number 5432, which you need to use later. Then install the graphical management UI pgadmin, which can be obtained from this link. After installation, log in to pgadmin and change the port in the server […]
-
Time：2020-5-22
Continued:Depth residual network + adaptive parameterized relu activation function (parameter adjustment record 7)https://blog.csdn.net/dangqin…In this paper, we test the effect of adaptive parametric relu (aprelu) activation function on the cifar10 image set. The keras code is as follows: #!/usr/bin/env python3 # -*- coding: utf-8 -*- “”” Created on Tue Apr 14 04:17:45 2020 Implemented using TensorFlow […]
-
Time：2020-5-20
In this paper, based on parameter record 9, shear is added to the data enhancement part_ Range = 30, test the effect of adaptive parametric relu (aprelu) activation function on cifar10 image set. For the usage of imagedatagenerator in keras, please refer to the following website:https://fairyonice.github.io/… Depth residual network + adaptive parameterized relu activation function […]
-
Time：2020-5-19
Some tutorials do not distinguish the difference between ordinary red black tree and left leaning red black tree. They directly teach the left leaning red black tree and call it red black tree, because compared with ordinary red black tree, the left leaning red black tree is easier to realize and teach. Here, we distinguish […]
-
Time：2020-5-19
In this paper, on the basis of parameter record 10, the number of residual modules is increased from 27 to 60. The effect of depth residual network using adaptive parametric relu (aprelu) activation function on cifar10 image set is tested. The keras program is as follows: #!/usr/bin/env python3 # -*- coding: utf-8 -*- “”” Created […]
-
Time：2020-5-19
Node connects to MySQL When it comes to node, you may think of mongodb as a database. Here we will introduce the connection between node and MYSQL, and share the encapsulated instance code, which can be used directly in project development. The next blog will talk about node connection to mongodb. Install MySQL module npm […]
-
Time：2020-5-19
This time, we try to solve the over fitting problem, reduce the number of residual modules to 2, reduce the weight of the first full connection layer in the adaptive parameterized relu activation function to 1 / 8 of the previous one, and set the batch size to 1000 (mainly to save time). The basic […]
