github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 16:11:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #14114] glm-ocr cannot process images > 2048x2048 or larger #71267

New Issue
Closed
opened 2026-05-05 01:00:17 -05:00 by GiteaMirror · 6 comments
GiteaMirror commented 2026-05-05 01:00:17 -05:00
Owner
Copy Link

Originally created by @slepkaviba on GitHub (Feb 6, 2026).
Original GitHub issue: https://github.com/ollama/ollama/issues/14114

Originally assigned to: @jmorganca on GitHub.

What is the issue?

When Given an image with Prompt "OCR" works fine if size is < 1800ish.

Here two images used for test - small one works, larger not.

Small:

Image

Large:

Image

Relevant log output

OS

Linux

GPU

RTX 4060Ti

CPU

5950X

Ollama version

0.15.5

Originally created by @slepkaviba on GitHub (Feb 6, 2026). Original GitHub issue: https://github.com/ollama/ollama/issues/14114 Originally assigned to: @jmorganca on GitHub. ### What is the issue? When Given an image with Prompt "OCR" works fine if size is < 1800ish. Here two images used for test - small one works, larger not. Small: <img width="925" height="1024" alt="Image" src="https://github.com/user-attachments/assets/f26c515e-41d4-4087-8f27-ba4b704b36ac" /> Large: <img width="1849" height="2048" alt="Image" src="https://github.com/user-attachments/assets/777d8b73-ee29-4f75-9282-9217df0dd850" /> ### Relevant log output ```shell ``` ### OS Linux ### GPU RTX 4060Ti ### CPU 5950X ### Ollama version 0.15.5
GiteaMirror added the bug label 2026-05-05 01:00:17 -05:00
GiteaMirror commented 2026-05-05 01:00:23 -05:00
Author
Owner
Copy Link

@letorbi commented on GitHub (Feb 6, 2026):

The problem does not seem to be the width or height of an image, but its total amount of pixels.

However, increasing the context size (https://github.com/zai-org/GLM-OCR/issues/32) works around the problem. I am using a context size of 10240 and have successfully extracted text from the example image from the GLM-OCR issue, which has 12 megapixels.

Here is the modelfile I am using:

# Modelfile generated by "ollama show"
FROM glm-ocr:latest

# Increase context size
PARAMETER num_ctx 10240

TEMPLATE {{ .Prompt }}
RENDERER glm-ocr
PARSER glm-ocr
PARAMETER temperature 0
<!-- gh-comment-id:3862232149 --> @letorbi commented on GitHub (Feb 6, 2026): The problem does not seem to be the width or height of an image, but its total amount of pixels. However, increasing the context size (https://github.com/zai-org/GLM-OCR/issues/32) works around the problem. I am using a context size of 10240 and have successfully extracted text from the example image from the GLM-OCR issue, which has 12 megapixels. Here is the modelfile I am using: ``` # Modelfile generated by "ollama show" FROM glm-ocr:latest # Increase context size PARAMETER num_ctx 10240 TEMPLATE {{ .Prompt }} RENDERER glm-ocr PARSER glm-ocr PARAMETER temperature 0 ```
GiteaMirror commented 2026-05-05 01:00:24 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Feb 18, 2026):

Confirm that increasing context lets glm-ocr handle larger images. I rendered a page of ipsum at 2500 dpi for an image size of 20671 x 29233 and glm-ocr with a context size of 8192 extracted the text accurately.

<!-- gh-comment-id:3921359876 --> @rick-github commented on GitHub (Feb 18, 2026): Confirm that increasing context lets glm-ocr handle larger images. I rendered a page of ipsum at 2500 dpi for an image size of 20671 x 29233 and glm-ocr with a context size of 8192 extracted the text accurately.
GiteaMirror commented 2026-05-05 01:00:27 -05:00
Author
Owner
Copy Link

@p-arndt commented on GitHub (Feb 26, 2026):

Somehow cant get it work even with bigger context size

<!-- gh-comment-id:3968905786 --> @p-arndt commented on GitHub (Feb 26, 2026): Somehow cant get it work even with bigger context size
GiteaMirror commented 2026-05-05 01:00:29 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Feb 26, 2026):

What version of ollama? 0.17.1 appears to have broken glm-ocr.

<!-- gh-comment-id:3968915772 --> @rick-github commented on GitHub (Feb 26, 2026): What version of ollama? 0.17.1 appears to have broken glm-ocr.
GiteaMirror commented 2026-05-05 01:00:31 -05:00
Author
Owner
Copy Link

@p-arndt commented on GitHub (Feb 26, 2026):

Ah okay that might explain it, i have ollama version is 0.17.1-rc1
Does version 0.17.0 work ? Version 0.17.0 works

<!-- gh-comment-id:3969076784 --> @p-arndt commented on GitHub (Feb 26, 2026): Ah okay that might explain it, i have `ollama version is 0.17.1-rc1` ~Does version 0.17.0 work ?~ Version 0.17.0 works
GiteaMirror commented 2026-05-05 01:00:32 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Feb 27, 2026):

Closing this for the size issue, the 0.17.1 breakage will be followed up in #14474.

<!-- gh-comment-id:3970111768 --> @rick-github commented on GitHub (Feb 27, 2026): Closing this for the size issue, the 0.17.1 breakage will be followed up in #14474.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label bug
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#71267
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?