Learning Rate Explorer

Understanding the Most Important Hyperparameter

Converging
Iteration: 0

Live Statistics

Loss 0.00000
Gradient Magnitude 0.000
Effective LR 0.000
Step Size 0.000
Loss History

Ready to Explore?

See how learning rate affects gradient descent. Too small creeps slowly. Too large explodes. Find the sweet spot.

💡 Drag to rotate • Scroll to zoom • Try the presets!