Error In Onehot: Depth Must Be >=2, But It Is 1

So I was working with ml5 in js on a simple NeuralNetwork that was going to be trained by images using the p5 js lib to get the images into a array and then train them through ml5,

Solution 1:

Not sure it is the same, but in my case some of the training datasets had only 1 single outcome of the output label, thus model would not train.

It seems to somehow match with the "depth must be >=2, but it is 1" message.

To be sure I just add another outcome manually in the dataset and relaunch