Skip to content
/ CNN-Numpy Public

Inplement an CNN frame with Numpy, easy to learn, hard to use hhhh

302 stars 65 forks Branches Tags Activity
Star
Notifications You must be signed in to change notification settings

wuziheng/CNN-Numpy

BranchesTags

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
.idea
.idea
 
 
data
data
 
 
layers
layers
 
 
old_layers
old_layers
 
 
README.md
README.md
 
 
conv_test.py
conv_test.py
 
 
lenet.py
lenet.py
 
 
old_lenet.py
old_lenet.py
 
 

Repository files navigation

CNN Inplementation With Numpy

近期准备找工作，感觉太久没有涉及深度学习相关的知识，面试起来十分生疏。准备简单的复习一下。现有的开源框架（tensorflow）对于小白使用起来抽象程度过高，易于使用，却不利于学习。所以我打算用numpy实现一个简单的可以训练，测试的cnn框架，包含一些主流的层和设计，以便复习与巩固基础。

2018.01.21

Target1: 实现mnist的训练与测试：

  • layer: Conv2D, FullyConnect, MaxPooling, Softmax
  • activation: Relu
  • method: Mini-batch Gradient Descent(SGD)

About

Inplement an CNN frame with Numpy, easy to learn, hard to use hhhh

Resources

Readme
Activity

Stars

302 stars

Watchers

13 watching

Forks

65 forks
Report repository

Releases

No releases published

Packages

No packages published

Languages