[Python] Facing a regenerate issue in my RAG application. When I select all sources, get a...

I'm using GPT35TURBO16K model, in my RAG application. I use Solr for doc search. The issue is that when I pass all sources, I get a shorter answer, but when I pass a subset of the sources, then I get a lengthy answer. What can be the reason for this? I tried changing stuff to map_reduce chain type for the answer generation. But still not able to figure out the exact cause of the issue.

Anyone up with a solution or approach to try out?

Expecting the regenerate issue to be resolved and I should get accurate answers based on provided sources only

Continue reading...