About RMSprop Optimizer Visualizer
This RMSprop visualizer runs the optimizer on Himmelblau's function, which has four separate minima, with a learning rate of 0.1 and the other three optimizers alongside. It is free, needs no sign-up and runs in your browser, so nothing you enter is uploaded.
RMSprop divides each coordinate's step by the root of a running average of its squared gradients. Directions with large gradients get smaller steps and directions with small gradients get larger ones, so it takes steps of roughly the learning rate in size almost regardless of the slope.
Himmelblau's function has four minima, all with value zero, so which one an optimizer reaches depends on its start and on how it moves. Tap different starting points and see where each of the four ends up.
How to use RMSprop Optimizer Visualizer
- Choose a function. The suggested learning rate for it is filled in.
- Change the learning rate if you like. Smaller is slower and safer, larger is faster until it blows up.
- Press Step to move every optimizer once, or Play to animate. "+25 steps" skips ahead.
- Tap the plot to move the starting point (the cross). The paths restart from there.
- Read the table for each optimizer's position, value and status: running, converged or diverged.
- Watch the log chart: a line that keeps falling is converging, and a line that climbs is diverging.
Features
- Seven functions: three curves and four 2D surfaces, including a narrow valley, Rosenbrock, Himmelblau and a saddle.
- SGD, Momentum, RMSprop and Adam, all from the same start with the same learning rate.
- Contour map for 2D functions and a curve for 1D, with a coloured path for each optimizer.
- Divergence detection that stops a run and says which optimizers blew up.
- A log-scale chart of the function value against step.
- Step, play and +25 steps controls, with slower playback if you prefer reduced motion.
- Tap the plot to choose the starting point.
RMSprop update
s is the running average of squared gradients, with decay 0.9.
s = 0.9 s + 0.1 g^2
x = x - lr g / (sqrt(s) + eps)Worked examples
Himmelblau from (0, 0), learning rate 0.1, 150 steps gives SGD diverged, Momentum diverged, RMSprop 0, Adam 0.0001
At this rate SGD and Momentum diverge, while RMSprop and Adam both settle into a minimum. From other starting points they can pick different ones of the four.
Privacy: how it works
- Everything is computed in your browser tab. Nothing is sent to Ulkaa or anyone else.
Tips
- On the parabola, try a learning rate of 0.5: SGD jumps to the minimum in a single step because the step size matches the curvature exactly.
- On the double well, drag the start to the left side and see it settle in the other, shallower minimum.
- On Himmelblau, change the start point and see different optimizers choose different minima.
- If a path vanishes off the plot, check the status column: it probably diverged.
Frequently asked questions
Is it really free?
Yes. There is no price, no trial and no credit card. Ulkaa Tool does the work with your own device, so there is no server bill to pass on to you.
Do I need an account?
No. There is no sign-up and no login. Open the page and use it. An Ulkaa account is not needed for anything on this tool.
Does it run in my browser?
Yes. Your settings are processed by code running in your browser tab, on your own device.
Is my data uploaded?
No. Your settings are never sent to Ulkaa or to anyone else. You can check this yourself by watching the Network tab in your browser developer tools while you use the tool.
What does RMSprop do?
It scales each parameter's step by the inverse root of a running average of its squared gradients, which evens out steps across directions of different steepness.
Is RMSprop the same as Adam without momentum?
Nearly. Adam adds a momentum term and bias correction to the RMSprop idea.
More Gradient Descent Visualizer pages
- Gradient Descent VisualizerCompare SGD, Momentum, RMSprop and Adam side by side, step by step, on curves and contour maps.
- Learning Rate Too High: Gradient Descent DivergenceSee the error grow instead of shrink when the learning rate crosses its limit.
- SGD Learning Rate VisualizerChange the learning rate and watch plain gradient descent speed up, oscillate or stall.
- Momentum vs SGDSee how a running velocity helps gradient descent in a long, narrow valley.
- Adam Optimizer VisualizerSee Adam adapt its step per direction on the curved Rosenbrock valley.
- Gradient Descent and Local MinimaStart on one side of a double well and see which minimum each optimizer reaches.
- Gradient Descent at a Saddle PointStart almost on a ridge and see how long each optimizer takes to escape the saddle.
- Gradient Descent on the Rosenbrock FunctionFollow gradient descent down the curved banana valley of the Rosenbrock function.
- Gradient Descent on a Bumpy LossSee how a bumpy loss traps plain gradient descent and how momentum helps.