Question about -prototype-aivtuber

[ponlponl123]

1. You can go to localhost:8000/tts to see the subtitle when running Node.
2. You can go to https://docs.aws.amazon.com/polly/latest/dg/voicelist.html to find a model for the language you want
3. You can tell the AI to add more about his story. Before asking your questions like:

await openai.createCompletion({
         model: 'davinci',
         prompt: `Hey CHARACTER_NAME! You are now CHARACTER_NAME who can answer anything. and is an expert. everyone will start talking Talk to you with the prefix Q:, but you don't need to prefix with the A:.\n
Q: YOUR_QUESTION
\nA:`,
         temperature: 0.9,
         max_tokens: 100,
         stop: ["ChatGPT:", "Adrian Twarog:", '\n'],
       });

[nakumiayumu144]

How can Ai read the names of people in chat?

[ponlponl123]

sure! You can use ${tags.username} to get the user's name!

[nakumiayumu144]

await openai.createCompletion({
         model: 'davinci',
         prompt: `Hey CHARACTER_NAME! You are now CHARACTER_NAME who can answer anything. and is an expert. everyone will start talking Talk to you with the prefix Q:, but you don't need to prefix with the A:.\n
Q: YOUR_QUESTION
\nA:`,
         temperature: 0.9,
         max_tokens: 100,
         stop: ["ChatGPT:", "Adrian Twarog:", '\n'],
       });

I'm using this code but I feel like the Ai is talking to itself.
i don't understand using model davinci

Something like that A: or Q:

[NovaPlayzGames]

Hey, firstly i would like to say using the normal davinci models are not recommended for chatting applications, this is basiclly what this project is, that being said i have made some adjustments on my repo to better reflect these changes and recommendations from open ai, https://github.com/InsanityLabs/AIVTuber