Do you have any errors in the control panel?Do you mean the user group permissions? They have all been set.
Then there are only two options: either the option to receive replies from AI is disabled in the threads, or the permissions are still not configuredNone whatsoever. Really clueless.
my users have been using it all day yesterday (although it is the day of least activity on my page) and they have only spent so far 0.10 USDAny idea what a typical Q&A costs for the api use? I'm thinking of putting this on our forum, in a dedicated area, but not if it is going to become a huge expense.
It looks like some really good coding on the XF side.
my users have been using it all day yesterday (although it is the day of least activity on my page) and they have only spent so far 0.10 USD
The update does not reset permissions. Permissions were only added in version 1.3.0, which is why you need to configure them after this update.Can you not upgrade without resetting all the permissions. IIt's a right PITA to have to redo them all.
That's right, in fact for now it only supports 3.5gpt turbo, but it's very good at what it does. I'm waiting for the addon that allows you to quote it in the commentsI assume your API is using gpt-3.5-turbo and not GPT-4?
no, chatgpt plus is for openai web. What this plugin uses is the API, they are two different thingsShould I upgrade to chatGPT plus? Any real advantages?
no, gpt plus chat is for openai web. What this plugin uses is the API, they are two different things
I'm expecting tremendous novelty usage from communities in the beginning. I'm curious what that engagement will look like in 1-2 years, and how communities with embedded AI bots will interact with the bots.I got this mod working roughly two hours ago. Members started figuring out how to play with it about one hour ago. Already 93 posts in the bot playpen and it's really just getting started. Fun stuff.
Hello. Yes, this functionality is planned for future updates.Would it be possible to add a configuration option for this mod consisting of an optional text box where we could define some text that would be injected with each request to ChatGPT (ie. appended to any post that is sent)?
Yes, the gpt-3.5-turbo model has a limit of 4096 tokens per request, you can read more about this model here:Question - does the mod, or the API, have a limit on the size of a query? Ie. how many characters can a member type into a post that ChatGPT will respond to?
We use essential cookies to make this site work, and optional cookies to enhance your experience.