run-llama/llama_cloud_services
1
0
Fork 0
mirror of https://github.com/run-llama/llama_cloud_services.git synced 2026-08-27 01:15:06 -04:00
Code Issues 352 Packages Projects Releases 10 Wiki Activity

Tables in markdowns are getting skewed #487

New Issue
Open
opened 2026-02-16 00:17:59 -05:00 by yindo · 3 comments
yindo commented 2026-02-16 00:17:59 -05:00
Owner
Copy Link
Copy Source

Originally created by @sowmyad on GitHub (Jun 19, 2025).

We are using llama-cloud with starter plan and wanted to explore PDF to markdown conversion for our GenAI usecase.

The PDFs that we use have heavy and complex tables and we noticed that llama-cloud is not able to convert the complex tables successfully in ALL cases. In some cases, it does. But for our use case, we will need conversion to be very high quality with 0 data loss.

Images can be ignored (and we don't need it). I am attaching some PDFs for reference. Several tables where there are colspans and rowspans, tables without headers are getting skewed.

Currently, we see that some pages and sections are getting skipped. To begin with look at how markdown is generated for RJTT. (AD 2.12).

AD 2-LFBO Characteristics.pdf
AD 2-LFPG Characteristics.pdf
AD 2-ZBAD Characteristics AMDB specific.pdf
RJTT AD 2 - Characteristics AMDB Specific.pdf

Originally created by @sowmyad on GitHub (Jun 19, 2025). We are using llama-cloud with starter plan and wanted to explore PDF to markdown conversion for our GenAI usecase. The PDFs that we use have heavy and complex tables and we noticed that llama-cloud is not able to convert the complex tables successfully in ALL cases. In some cases, it does. But for our use case, we will need conversion to be very high quality with 0 data loss. Images can be ignored (and we don't need it). I am attaching some PDFs for reference. Several tables where there are colspans and rowspans, tables without headers are getting skewed. Currently, we see that some pages and sections are getting skipped. To begin with look at how markdown is generated for RJTT. (AD 2.12). [AD 2-LFBO Characteristics.pdf](https://github.com/user-attachments/files/20809878/AD.2-LFBO.Characteristics.pdf) [AD 2-LFPG Characteristics.pdf](https://github.com/user-attachments/files/20809881/AD.2-LFPG.Characteristics.pdf) [AD 2-ZBAD Characteristics AMDB specific.pdf](https://github.com/user-attachments/files/20809880/AD.2-ZBAD.Characteristics.AMDB.specific.pdf) [RJTT AD 2 - Characteristics AMDB Specific.pdf](https://github.com/user-attachments/files/20809879/RJTT.AD.2.-.Characteristics.AMDB.Specific.pdf)
yindo commented 2026-02-16 00:18:00 -05:00
Author
Owner
Copy Link
Copy Source

@hexapode commented on GitHub (Jun 19, 2025):

Thanks for reporting, having a look at them

@hexapode commented on GitHub (Jun 19, 2025): Thanks for reporting, having a look at them
yindo commented 2026-02-16 00:18:00 -05:00
Author
Owner
Copy Link
Copy Source

@sowmyad commented on GitHub (Jun 19, 2025):

RJTT_39.pdf

This page has one good example where the table was completely skewed in markdown. (It is one of the page from earlier uploaded PDF). Tried converting PDF -> DOCX file and uploaded DOCX file to llama-cloud as well, just to see the performance. The table was still skewed.

@sowmyad commented on GitHub (Jun 19, 2025): [RJTT_39.pdf](https://github.com/user-attachments/files/20814271/RJTT_39.pdf) This page has one good example where the table was completely skewed in markdown. (It is one of the page from earlier uploaded PDF). Tried converting PDF -> DOCX file and uploaded DOCX file to llama-cloud as well, just to see the performance. The table was still skewed.
yindo commented 2026-02-16 00:18:00 -05:00
Author
Owner
Copy Link
Copy Source

@sowmyad commented on GitHub (Jul 3, 2025):

Hi,
Wanted to check if there is an update on this issue?

@sowmyad commented on GitHub (Jul 3, 2025): Hi, Wanted to check if there is an update on this issue?
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#487
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?