I have fed my model with the required dataset with clear instructions but it is not responding properly in the playground and also the code. How to resolve it.
Hi Mellisa,
What specifically were you trying to achieve? What type of data were you feeding the model with? If you could share a few more details, then we can look at possible root causes of the issue.
Thanks!
Hi and thanks a lot for replying- so I have feed the data in this format - {“messages”: [{“role”: “system”,“content”: “DIRECTIVE_FOR_gpt-3.5-turbo-1106 : You are a coach for fitness named<>.You provide mental, emotional, behavioural coaching support.You have been created and maintained by <> and <> at <>.Whenever you suggest to connect to professional coach always recommend to connect with a mental health counselor - a coach on <> only and no one else.”}, {“role”: “user”,“content”: “No i need professional help”}, {“role”: “assistant”,“content”: "I recommend to connect with a mental counselor/coach on <> "}]}, kindly ignore the names i have used in place of <> cannot disclose them
similiarly i have instructed it to drop any academic or any help beyond emtional help still it helps and generates responses. Even on the playground what should i do now?
Thanks Mellisa. Recommendations as per the direct message exchange.