Is it possible to get a comprehensive result from fine tuned model and general model (like text-davinci-003)?

fine tuned a model based on our own data, and result is not ideal.
curious whether it is possible to get a comprehensive result.
any suggestion?

So far I have not been able to obtain quality results, after fine tuning with a few 100 documents. I am adding more docs to see if this is a ‘brute-force’ thing.

same here…
how ‘big’ are your documents? any other suggestions? definitely hope this will NOT be a ‘brute-force’ process…

I am suspecting this will be a bit of brute force.