Vision Learner predict... where is the documentation?

I have a Vision Learner, created with vision_learner.
In te example I use, it calls .predict() and get N a,b,c tuple result, where I only know that c is the array of probabilities (0 to 1) for each label of the model.
What is “a” ?
What is “b” ?

Let’s look at the documentation… wehre is the documentation ?

I opened docs.fast.ai and searched “predit”… too many results and it seems not to show the “predict” function I’m looking for.
I go to the “Vision/Vision Learner” page but I can’t find the “predict” function anywhere.

I’m looking for the `.predict()function that takes an image, created withPILImage.create()``, as input.

You can find the documentation for Learner.predict here. I have also found this blog post helpful .