Skip to content

mtmd: add mtmd_get_vision_image_size() and mtmd_get_vision_patch_size() functions #16705

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 1 commit into from
Closed

mtmd: add mtmd_get_vision_image_size() and mtmd_get_vision_patch_size() functions #16705

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
16 changes: 16 additions & 0 deletions tools/mtmd/mtmd.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -834,6 +834,22 @@ int mtmd_get_audio_bitrate(mtmd_context * ctx) {
return 16000; // 16kHz
}

int mtmd_get_vision_image_size(mtmd_context * ctx) {
if (!ctx->ctx_v) {
return -1;
}

return clip_get_image_size(ctx->ctx_v);
}

int mtmd_get_vision_patch_size(mtmd_context * ctx) {
if (!ctx->ctx_v) {
return -1;
}

return clip_get_patch_size(ctx->ctx_v);
}

//
// public API functions
//
Expand Down
8 changes: 8 additions & 0 deletions tools/mtmd/mtmd.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -112,6 +112,14 @@ MTMD_API bool mtmd_support_audio(mtmd_context * ctx);
// return -1 if audio is not supported
MTMD_API int mtmd_get_audio_bitrate(mtmd_context * ctx);

// get vision image size in pixels, for example 1024
// return -1 if vision is not supported
MTMD_API int mtmd_get_vision_image_size(mtmd_context * ctx);
Comment on lines +115 to +117
Copy link
Collaborator

@ngxson ngxson Oct 22, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Not a good idea to add this, as many models now support dynamic resolution and this won't give the correct size

Copy link
Contributor Author

@deadprogram deadprogram Oct 22, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Indeed, some models are fixed res, some are naflex.

The presumption is that if you want to do your own pre-processing, you will need to be familiar with what the model you are using expects.


// get vision patch size, for example 14
// return -1 if vision is not supported
MTMD_API int mtmd_get_vision_patch_size(mtmd_context * ctx);

// mtmd_bitmap
//
// if bitmap is image:
Expand Down
Loading