ENNUI
ENNUI copied to clipboard
An Elegant Neural Network User Interface to build drag-and-drop neural networks, train in the browser, visualize during training, and export to Python.
ENNUI
ENNUI is an Elegant Neural Network User Interface that allows users to:
- Build neural network architectures with a drag and drop interface.
- Train those networks on the browser.
- Visualize the training process.
- Export to Python.
Thanks to Gil Strang, it's hosted at https://math.mit.edu/ENNUI/

Installation
After cloning the repo, to install dependencies, run:
npm install
To Build the app, run:
npm run build
For Continuous Build and Integration, run:
npm run watch
To use chrome with localhost and the CIFAR-10 Dataset, make sure to close all chrome windows and run the following command to launch chrome: 'path/to/chrome.exe --allow-file-access-from-files'
Code Structure
ENNUI is frontend-only. The implementation is structed in two parts:
-
ENNUI/src/ui
is home to the traditional frontend components such as the styling, the button logic, and the draggable components (e.g. the layers and activations). -
ENNUI/src/model
is the "backend" implementation supporting core functionality such as building neural networks, doing the code generation, and saving the state in the browser.