This version is for ServUO 58 : I might do a version for 57, just won't be soon as I am currently working on another project!

Also you need to get your own key and edit it into the script, not sure you did that, otherwise the rest is due to Net Version issues that can be fixed but like I said, I don't have the time atm to revisit this!
 
Thanks, all right, I'll test it on the '58 version, by the way, the design of the black box looks good. (Jedi66)

As far as that works, only Tom is missing, is not found.

It's a pity that there isn't a complete 58 version. (Scripts)

CamBild135861.png
 
Last edited:
Thanks for making this it is something I always wanted :D I did have a problem with the NPC just saying he needs time to think in various ways instead of responding in any meaningful way. Do you know why that would happen?
 

Attachments

  • Screenshot 2023-04-06 194147.png
    Screenshot 2023-04-06 194147.png
    309.4 KB · Views: 10
@Nikodemus You can clone the github for the most recent 58, and with 58 you must recompile each time to add new scripts, different than 57 that does it on launch every time regardless!

@Izex The way I have the AI intergraded and to make sure there is no waiting for a response from the API, there is a storage que and such but basically the responses are there and will filter in as you use it! It also prevents abuse as the system is a pay per use. So after the same question is asked 10 times and there are 10 replies from the ai for that question, any future questions are answer locally by the storage and not the ai, which saves on usage! If you read up the thread, I covered this in detail with another member!
 
Thanks for the info. Has anyone here already compiled the 57 with the 58 version? That's the next thing I'll do then
 
57 and 58 are night and day, it is like you just said lets compile windows 11 into windows 8! You either use 57 or use 58, or any of the previous versions from 56, 55, 54, RunUO 2.2, 2.1, 2.0 or 1.0... I mean since RunUO started back in 2004, we have been continuously moving forward! I usually code for the latest but sometimes will convert systems to work on earlier versions!
 
The problem is that the AI never responded properly even once. It might have been my API key then, thanks for the answer.
 
Hello
I need your help
Tell me what am I doing wrong?
api key added
account starting 5$
04:13:10 Error calling OpenAi API to get completion. HTTP status code: NotFound. Request body: {"prompt":"Thou art now to act the role of a Confident Bard NPC in the fantastical realm of Ultima Online. Speak thy response in a single, grandiose sentence, if thou wouldst oblige: hi.\n\n ","max_tokens":1024,"temperature":0.5,"top_p":0.3,"presence_penalty":0.1,"frequency_penalty":0.1,"n":1,"stream":false,"logprobs":0, "echo":false,"stop":"."}
04:13:13 Error calling OpenAi API to get completion. HTTP status code: NotFound. Request body: {"prompt":"Thou art now to act the role of a Mad Bard NPC in the fantastical realm of Ultima Online. Speak thy response in a single, grandiose sentence, if thou wouldst oblige: hi.\n\n ","max_tokens":1024,"temperature":0.5,"top_p":0.3,"presence_penalty":0.1,"frequency_penalty":0.1,"n":1,"stream":false,"logprobs":0, "echo":false,"stop":"."}
 
I'll look into this! Things might have changed since I last visited this script and OpenAI API. won't be anything major, just a bad call to the API, probably the model as this was made when Davinci was top model to pick from, now we can access gpt4!
 
I'll look into this! Things might have changed since I last visited this script and OpenAI API. won't be anything major, just a bad call to the API, probably the model as this was made when Davinci was top model to pick from, now we can access gpt4!
thank you for your answer, I hope for your support
NPC does not respond to queries in the log, these are the errors all the time
 
After reviewing the changes to the models and openai api, and some thought, I decided this resource will only be for educational reference only and not updated!

Added to my todo list, create a new system based on the new models/api.

With everything I have now, I estimate about a week to complete and uploaded for those interested!
 
After reviewing the changes to the models and openai api, and some thought, I decided this resource will only be for educational reference only and not updated!

Added to my todo list, create a new system based on the new models/api.

With everything I have now, I estimate about a week to complete and uploaded for those interested!
this is great news
 
Nope, this will be the OpenAI Version, if there is demand, I'll do one for GTP4ALL < Just not as good so not sure if a good choice tbh!
Progress Update, I might have this out later tonight!

View attachment 23357View attachment 23358
View attachment 23359
Great
you're right gtp4all is not so good but it seems suitable for simple dialogues, maybe it is possible to choose between gpt and gpt4all by switching settings or changing the config
 
Great
you're right gtp4all is not so good but it seems suitable for simple dialogues, maybe it is possible to choose between gpt and gpt4all by switching settings or changing the config
Its not hard plugging in the GPT4ALL, it comes with its own API Service, so just a matter of pointing to it! I'll probably add it when I do the UI for the UO OpenAI Server! For now I just want to make sure the main system is solid before expanding!
 
Back