Newbie question: Why is "learn sound model" greyed out when trying to remove background sounds in a file using the Spectral Frequency display?

I am trying to remove background sounds in a file using the Spectral Frequency display. The first edit on a new file, this process works fine:

1. use brush to cover sound

2. right click, learn sound model

3. effects/noise-reduction-restoration/Sound Remover

The next time I try to remove another sound in the file, "learn sound model" is greyed out. I have to close the file and reopen to be able to make another edit. What on earth is going on? Can it only learn one sound model or something?

I'm a newbie, so thanks in advance for your advice!