News for hackers. Stuff that matters.
Re:Show HN: IncarnaMind-Chat with your multiple docs using LLMs by gsuuon on Friday September 15, 2023 @07:51PM (#37530457)
So to rephrase: stage 1 - it finds the top k most relevant small chunks, stage 2 - it searches the source documents of those small chunks for most relevant large chunks, stage 3 - it searches medium chunks in the source documents of relevant small chunks that are contained in the large chunks found in stage 2?
Re:Show HN: IncarnaMind-Chat with your multiple docs using LLMs by joeyxiong on Friday September 15, 2023 @08:09PM (#37530602)
Yeah, the sliding window chunking mechanism appears in stage 2.