Boadzie / kit-toxic

This is a simple web app that uses TensorFlow.js and a pre-trained toxicity model to predict whether a given sentence is toxic or not and the type of toxicity

Detecting Online Toxicity with Tensorflow.js

This is a simple web app that uses TensorFlow.js and a pre-trained toxicity model to predict whether a given sentence is toxic or not. The app is built with SvelteKit and styled with Tailwind CSS, and features a loading spinner, a fly animation, and a CSS animation using keyframes.

How It Works

The app consists of a single page (+page.svelte) with a form that takes a sentence as input and a "Check" button to submit it. When the user clicks the button, the app loads the toxicity model, classifies the sentence, and displays the results in a scrollable panel.

The app also features a loading spinner while the model is being loaded, a fly animation when the results are displayed, and a CSS animation that makes the title pulse continuously.

Developing

Once you've forked, cloned the project and installed dependencies with npm install (or pnpm install or yarn), start a development server:

pnpm run dev

Building

To create a production version of your app:

pnpm run build

You can preview the production build with npm run preview.

To deploy your app, you may need to install an adapter for your target environment.