Skip to content

Python tool that uses GTK3 to see the calculatiosn in small torch neural networks.

License

Notifications You must be signed in to change notification settings

mrquorr/nn_visualizer

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

nn_visualizer

Python tool that uses GTK3 to see the calculatiosn in small torch neural networks.

Limitations

A limit of 40 neurons was added to ensure visualizations aren't too large.

Installation

Make sure you have the packages installed.

pip install -r requirements.txt

Known issues

I had some issues installing PyGObject because gi library 2.0 is not available in the latest stable debian release, so I had to install a PyGObject compatible with gi library 1.0.

pip install PyGObject==3.46.0

Visualizer description

v3 shows a detailed breakdown of calculations on each layer

Running it

you can run with python examples.py, this will run the demonstration with the 5 examples listed there. You can modify the examples for different visualizations, or add your own.

Forward propagation example

Backwards propagation example

About

Python tool that uses GTK3 to see the calculatiosn in small torch neural networks.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages