You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm extracting information from bank account statement PDFs, that are sometimes over 10 pages long. Passing all of the markdown content to an LLM is not ideal because of the context length, so my solution is chunking.
However as I'm looking into bank account statements I cannot have a transaction line items cut in half (imagine if what gets cut is the transaction value and what it was for).
The model output (at least in markdown) doesn't seem to include page number ("page 1", "page 2"). Is there a way that the model can be forced to do that? Or another technique that is not what I'm describing but could still solve my problem?
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
I'm extracting information from bank account statement PDFs, that are sometimes over 10 pages long. Passing all of the markdown content to an LLM is not ideal because of the context length, so my solution is chunking.
However as I'm looking into bank account statements I cannot have a transaction line items cut in half (imagine if what gets cut is the transaction value and what it was for).
The model output (at least in markdown) doesn't seem to include page number ("page 1", "page 2"). Is there a way that the model can be forced to do that? Or another technique that is not what I'm describing but could still solve my problem?
Beta Was this translation helpful? Give feedback.
All reactions