Skip to content

jaebooker/compressed_cnn

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Install requirements Run "python cnn.py"

Arhictecture: Three convolutional layers with batchnorm and maxpooling applied Four hidden FCN layers with relu activation Softmax output layer

Uniform quanitzation to 8 and 6 bits

Unstructured pruning below a given threshold

Student-Teacher Knowledge Distillation

About

Using compression methods on a CNN, including Pruning, Quantization, and Student-Teacher Knowledge Distillation

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages