Skip to content

Re-enable upscaling of images smaller than the CLIP input size; fix MiniCPM evaluation on small bitmaps#13237

Merged
ngxson merged 1 commit into
ggml-org:masterfrom
lcarrere:master
May 1, 2025
Merged

Re-enable upscaling of images smaller than the CLIP input size; fix MiniCPM evaluation on small bitmaps#13237
ngxson merged 1 commit into
ggml-org:masterfrom
lcarrere:master

Conversation

@lcarrere

@lcarrere lcarrere commented May 1, 2025

Copy link
Copy Markdown
Contributor

Since commit #13011, MiniCPM models have been unable to process small images. This appears to be caused by disabling the upscaling of images smaller than the CLIP input size. This PR reintroduces that upscaling and restores full functionality for these models.

I'm attaching 2 example of small image for reference
10
cat0

@ngxson
ngxson merged commit b6e4ff6 into ggml-org:master May 1, 2025
timwu pushed a commit to timwu/llama.cpp that referenced this pull request Dec 20, 2025
Seunghhon pushed a commit to Seunghhon/llama.cpp that referenced this pull request Apr 26, 2026
ljubomirj pushed a commit to ljubomirj/llama.cpp that referenced this pull request May 6, 2026
my-other-github-account pushed a commit to my-other-github-account/llama.cpp that referenced this pull request May 15, 2026
my-other-github-account pushed a commit to my-other-github-account/llama.cpp that referenced this pull request May 15, 2026
phibya pushed a commit to ziee-ai/llama.cpp that referenced this pull request May 29, 2026
AlexiAlp pushed a commit to minghaop/llama.cpp that referenced this pull request Jun 2, 2026
AlexiAlp pushed a commit to minghaop/llama.cpp that referenced this pull request Jun 2, 2026
fukuro-kun pushed a commit to fukuro-kun/fukuro-llama-cpp-turboquant that referenced this pull request Jul 5, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants