Training details
#2
by
tobiasnorlund
- opened
Hi! How did you format the prompts when training this model? Did you simply use the name field, e.g. detect {name} which would be detect share screen button for this example? Did you also train using other fields such as description?
That's right, only detect {name}.