The annotators have to make annotations for several labels (REVIEW, DEAL) etc. in the binary textcat.manual and textcat.teach modes (and different data). For efficiency it's easiest that the annotators can make annotations for several labels without me changing the command when they wish to make a new set of annotations.
Anyone had a similar problem, and how did you solve that?
Hi! I hope I understand your question correctly, but you could just have two processes running on different ports on the same machine? Also see my comment on your other question here:
If you're annotating with a model in the loop, your process will be inherently stateful so you definitely want to have separate isolated processes here. And then you can have your annotators connect to either of them, depending on what they want to annotate.