Its a compound problem. The client isn’t sending something the server needs to setup the conversation. So the server doesn’t respoyin kind unfwwe just haven’t figured out what that thing is yet.
Im pretty sure it’s in the integration as the session is getting created. If the experiment @mirbusak does yields fruit I’d suggest opening a github issue with that information. They’ll probably be able to figure out what’s going on with what you’ve collected so far.
I’m not using the recommended model settings, I already have that turned off and I’ve tried changing the models, still the same result. I’m giving up and will settle for the extended open AI for now, even though it’s a lot slower.
I enabled another model, the exact one the log message referred to, and still the same result, I also tried gpt 3.5 turbo, the same thing,
my allowed models:
Now I don’t know exactly what you want from me?
I opened the link, but what about it, I don’t speak English and I translate everything and that’s a big problem.