Create an image

Given a cue, the model will return one or more predicted completions, and may also return probabilities of alternative labels for each location.

Last updated