Not at the moment, I am not against adding some sort of outlet here so you can style it better, but for now there is no way.
Some Discourse instances are setup as help resources and moderators are the ones often answering many questions as humans, despite search and FAQ facilities. Could there be a way for an ai persona to be styled on the basis of such a moderatorās historical interactions? So train the model on one userās tone of voice, vocabulary and other syntactical aspects?
This could greatly assist the moderatorās job as the ai persona would be replying in a style very familiar to a longstanding existing user base.
Persona/bot seems to be working sporadically for me when mentioned. In the same topic it sometimes work, sometimes not. I double checked and all permissions (allowed groups in persona settings, category permissions for bot to reply) seem to be set correctly. What is the proper way to debug it? No error in /logs.
Iām having the same problem, can you please tell me if youāve solved it, thanks!
I donāt speak English very well, this is a machine translation, so please forgive me if Iām offended.