run-llama/llama_cloud_services
1
0
Fork 0
mirror of https://github.com/run-llama/llama_cloud_services.git synced 2026-08-26 16:56:36 -04:00
Code Issues 352 Packages Projects Releases 10 Wiki Activity

LLamaparse is missing chunks of text when parsing PDF / How Do You Ensure Your Parser Fully Parses a Document Without Missing Content (Text/Tables/Information)? #392

New Issue
Open
opened 2026-02-16 00:17:42 -05:00 by yindo · 3 comments
yindo commented 2026-02-16 00:17:42 -05:00
Owner
Copy Link
Copy Source

Originally created by @MuhammedTech on GitHub (Dec 24, 2024).

I’ve been testing LlamaParse for PDF parsing, and I was surprised to find that when I manually checked the output, some text seemed to be missing. I’m wondering how others ensure that the parser truly processes the entire document and doesn't leave out or miss any important pieces of information (text, tables, etc.).

How do you guys test your parsers to make sure they parse the whole document without any omissions? Do you use any specific validation techniques or post-processing checks to ensure completeness?

I’d love to hear your experiences and recommendations for improving document parsing accuracy

Originally created by @MuhammedTech on GitHub (Dec 24, 2024). I’ve been testing LlamaParse for PDF parsing, and I was surprised to find that when I manually checked the output, some text seemed to be missing. I’m wondering how others ensure that the parser truly processes the entire document and doesn't leave out or miss any important pieces of information (text, tables, etc.). How do you guys test your parsers to make sure they parse the whole document without any omissions? Do you use any specific validation techniques or post-processing checks to ensure completeness? I’d love to hear your experiences and recommendations for improving document parsing accuracy
yindo commented 2026-02-16 00:17:43 -05:00
Author
Owner
Copy Link
Copy Source

@galvangoh commented on GitHub (Dec 29, 2024):

I am interested to understand this as well. From my testing, I realized long and complicated parsing instructions tends to degrade the quality of output (e.g. table is being parsed but the contents within are being reduced, as though summarized to a minimal).

@galvangoh commented on GitHub (Dec 29, 2024): I am interested to understand this as well. From my testing, I realized long and complicated parsing instructions tends to degrade the quality of output (e.g. table is being parsed but the contents within are being reduced, as though summarized to a minimal).
yindo commented 2026-02-16 00:17:43 -05:00
Author
Owner
Copy Link
Copy Source

@rthomas67 commented on GitHub (Jan 3, 2025):

I had the same experience with a document-only / default-params call to the parsing/upload REST API endpoint, as described here, but there are quite a few parameters (documented here) that might help optimize the completeness of the output. If I find out anything helpful, I'll add another comment here with details.

@rthomas67 commented on GitHub (Jan 3, 2025): I had the same experience with a document-only / default-params call to the `parsing/upload` REST API endpoint, as described [here](https://docs.cloud.llamaindex.ai/llamaparse/getting_started/api), but there are quite a few parameters (documented [here](https://docs.cloud.llamaindex.ai/API/upload-file-api-v-1-parsing-upload-post)) that **might** help optimize the completeness of the output. If I find out anything helpful, I'll add another comment here with details.
yindo commented 2026-02-16 00:17:43 -05:00
Author
Owner
Copy Link
Copy Source

@judithnat commented on GitHub (Feb 17, 2025):

I have found llamaparse misses out some or all information on later pages. Anyone know of the maximum number of pages llamaparse is supposed to be reliable for?

@judithnat commented on GitHub (Feb 17, 2025): I have found llamaparse misses out some or all information on later pages. Anyone know of the maximum number of pages llamaparse is supposed to be reliable for?
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
LlamaCloud
LlamaExtract
LlamaParse
bug
client
complex layout
documentation
duplicate
enhancement
figure
font
header footer
help wanted
ocr
parser
pdf_debug
pdf_debug
pull-request
Mirrored from GitHub Pull Request
 python-lib
question
server_issue
server_issue
ui
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: run-llama/llama_cloud_services#392
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?