I'm trying to use GPT-3.5 turbo for text summarization of large transcriptions, However i'm running up against the 4096 token limit. Is there a better way to do this that i'm not thinking of?
#Larger inputs than the maximum number of tokens
polar hatch
cinder jasper
Chunk the text into what will fit in 3500 tokens and allow a 590 token response. Then do it again, then merge the responses.
hasty inlet
also have a look at the embedding API
polar hatch
The embedding api is very Interesting however I’m not sure it’s useful for what I’m trying to do. I’m not trying to model similarities or perform search, just take in a large human readable document and summarize it into something more condensed but still meant for humans
polar hatch
Maybe I’m just confused though
polar hatch
Yeah I don’t think most of this will work since I want a human readable summary, although I do have some ideas.