Skip to content

Carrot is a PyTorch neural network analysis tool using Dash.

Notifications You must be signed in to change notification settings

andreiliphd/carrot

Repository files navigation

Carrot - tensorboard for PyTorch

============

Carrot is a web application that uses MySQL to get data of loss, accuracy, optimized parameters and gradients that allows to visually see what is happening inside a neural network.


Features

  • Visually see gradients and parameters in development
  • Analyse losses and accuracy

Screenshot

Carrot - gradient and parameters analysis tool for PyTorch


Setup

Clone this repo:

git clone https://github.com/andreiliphd/carrot.git

Install all the dependencies.


Usage

  • You can use this application to analyse gradients and parameters visually
  • You can use Carrot for educational purposes

History

  • It used to save everything in MongoDB
  • Now MySQL is used
  • Fixed problem with speed of loading

Contributing

  • Sometimes UI do not refresh possibly problem with callbacks

License

You can check out the full license in the LICENSE file.

This project is licensed under the terms of the MIT license.

About

Carrot is a PyTorch neural network analysis tool using Dash.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published