AI Studio deployment using pre-built knowledge base and custom knowledge base simultaneously

Question

Working with AI Studio and created a GPT 3.5 deployment. Tested deployment by typing any random physical location. It is able to provide description and even directions which I assume, are derived from pre-built knowledge base.

When I add my own custom knowledge, it is able to answer queries described in my knowledge base, but no longer returns feedback for generic queries such as the one described above.

Is it possible for the deployment to use both my knowledge as well as the pre-built knowledge simultaneously?

Answer 1

Hello,

Thanks for reaching out to us, I think there may be a misunderstanding for the feature Add Your Own Data and the feature Fine-tuning.

Add your own data allows for fine-grained customization and domain-specific adaptation, while fine-tuning is closer than your user case, which leverages OpenAI's pre-trained models and Azure's infrastructure for efficient training,

Please refer to the document here - https://learn.microsoft.com/en-us/azure/ai-services/openai/how-to/fine-tuning?tabs=turbo%2Cpython-new&pivots=programming-language-studio

I hope this helps!

Regards,

Yutong

Answer 2

You are correct, went through documentation again (slowly), adding own data will utilize that new dataset exclusively. Was not aware of 'fine-tuning' but seems to fit my description, will need to read up on that.