Teaching kids deep learning with intuitive drag-n-drop block-based interface.

View on Github
Developed in the Pytorch Summer Hackathon 2019
View on Github
3rd Place Winner of Pytorch Summer Hackathon 2019

Demo

Click Play Button to Demo

Features

Augmented Deep Learning
Components
Custom blockly components: convolutional, fully connected, dataset switcher, GuessCategory,
'improve model with data'
one click optimizer.
Drag & Drop
Layout
Kid-friendly, colorful, intuitive
Drag-n-Drop block-based
Drag to add and
remove NN layers
Web Based
Pure Js
Computer agnostic, light-weight browser based interace.
Can run on any browser,
embed on any web page.

Team

Erin Song
Facebook Udacity Private Secure AI Scholar, Proud Mother of Eli (who also contributed to testing), Educator, USA-based Data Scientist
Anna Unger
Facebook Udacity Private Secure AI Scholar, Mother of four, Swedish Data Scientist and Lifetime Learner. Facebook Developer Circle.
Boyuan Chen
AI Researcher at Berkeley. USA-based. Robotics enthusiast. Facebook Developer Circle.
Yu Sun
Current an ML mentor at a leading edtech startup. USA-based.