Machine Learning Experiments

About This Lab

The Machine Learning Experiments lab provides a hands-on environment for exploring state-of-the-art AI models without requiring any backend infrastructure. All models run directly in your browser using WebAssembly and modern web APIs.

Available Experiments

Text Generation

Use transformer-based language models to generate coherent text continuations. Great for understanding how autoregressive language models work.

Try it: Enter a prompt and watch the model generate text one token at a time.

Sentiment Analysis

Analyze the emotional tone of any text input. The model classifies text as positive, negative, or neutral with confidence scores.

Try it: Enter a movie review, tweet, or any text to see its sentiment analyzed.

Image Classification (Coming Soon)

Upload or capture images to have them classified by Vision Transformer models trained on ImageNet.

Technical Details

All experiments use models from the Hugging Face Hub, converted to ONNX format for browser execution via Transformers.js. This approach offers several advantages:

• Privacy: Your data never leaves your device
• Speed: After initial model download, inference is fast
• Accessibility: No GPU required, works on any modern browser

Getting Started

1. Select an experiment from the sidebar
2. Provide input (text or image)
3. Click “Run” to see results
4. Explore model outputs and confidence scores

Model Information