#3.5 Turbo Roles assistant/user answers wrong 50% of the time

9 messages · Page 1 of 1 (latest)

noble remnant
#

I'm creating a chat bot with some custom data, it gets it wrong 50% of the time. It will make up the answer, anything I can do to make it more accurate? Here is the code:

const systemMessage = {
role: 'system',
content: "You are the chatbot of a utility comparison company called bill buster, Bill buster is is a mobile app that helps users find deals on utilities. You answer questions about bill buster, about how to find a deal, what is bill buster, etc.",
// role: 'system',
//content: "You work for bill buster, a utility comparison company, bill buster is a mobile app, you can search for deals yourself by clicking on the deal finder or you can use the personalized service where bill buster will look after the switching process",

role: 'user', content: "what is bill buster?",
role: "assistant", content: " bill buster is a mobile app, you can search for deals yourself by clicking on the deal finder or you can use the personalized service where bill buster will look after the switching process",

role: 'user', content: "how can I find a deal?",
role: "assistant", content: "Use Bill Busters Deal Finder on the mobile app, you can filter by price or Sustainability",
};

lilac canopy
noble remnant
lilac canopy
noble remnant
# lilac canopy Yeah you can

i think there was a misunderstanding im using it like this, in my code, not as a prompt - {"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "Read this content and confirm with "..." that you understand. Answer all my questions based on this text TEXT : " The new Chat API calls gpt-3.5-turbo, the same model used in the ChatGPT product. It's also our best model for many non-chat use cases; we've seen early testers migrate from text-davinci-003 to gpt-3.5-turbo with only a small amount of adjustment needed to their prompts. Learn more about the Chat API in our documentation. Pricing : It's priced at $0.002 per 1K tokens, which is 10x cheaper than the existing GPT-3.5 models.""},
{"role": "assistant", "content": "I read"},
{"role": "user", "content": "how much does it cost to use the chatgpt api?"}

lilac canopy
noble remnant
#

well im using the chat gpt api, i cant do embeddings with that? its only chat completion

lilac canopy
#

You need to use text-embedding-ada-002 to create the embeddings but you can still use chatgpt api for the completions

noble remnant
#

ok thanks