parsing glossary across pages in pdf breaks up definition #211

Open
opened 2026-02-16 00:17:09 -05:00 by yindo · 0 comments
Owner

Originally created by @gitcagey on GitHub (Jul 26, 2024).

Originally assigned to: @BinaryBrain on GitHub.

using llamaparse api to parse a pdf into text.

take this pdf: https://www.aphis.usda.gov/sites/default/files/hawaii.pdf

this definition is across 2 pages in pdf:
image

but when parsed it looks like this:
image

this is problematic as using gemini 1.5 pro to provide me the glossary terms and results in a RECITATION stop reason in their API. Bottom line is if I fix the broken parsed term/definition manually I don't get the problem in GEMINI. But the issue is invalid parsed text out of llamaparse.

Originally created by @gitcagey on GitHub (Jul 26, 2024). Originally assigned to: @BinaryBrain on GitHub. using llamaparse api to parse a pdf into text. take this pdf: https://www.aphis.usda.gov/sites/default/files/hawaii.pdf this definition is across 2 pages in pdf: ![image](https://github.com/user-attachments/assets/da265ed2-30df-4286-a708-11f4f404ccac) but when parsed it looks like this: ![image](https://github.com/user-attachments/assets/62c16821-6ab8-4f99-8881-b808ed1a9b36) this is problematic as using gemini 1.5 pro to provide me the glossary terms and results in a RECITATION stop reason in their API. Bottom line is if I fix the broken parsed term/definition manually I don't get the problem in GEMINI. But the issue is invalid parsed text out of llamaparse.
yindo added the bug label 2026-02-16 00:17:09 -05:00
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: run-llama/llama_cloud_services#211