Skip to content

fix(mmmu_pro_vision): apply post_prompt to vision-only split #1336

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

Merged
kcz358 merged 1 commit into from
May 20, 2026
+1 −1
Merged

fix(mmmu_pro_vision): apply post_prompt to vision-only split #1336

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
2 changes: 1 addition & 1 deletion lmms_eval/tasks/mmmu_pro/utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -52,7 +52,7 @@ def mmmu_pro_doc_to_text(doc, lmms_eval_specific_kwargs=None):
return mmmu_doc_to_text_qwen3vl(doc, lmms_eval_specific_kwargs)

post_prompt = lmms_eval_specific_kwargs["post_prompt"]
question = doc.get("question", "")
question = post_prompt # vision-only fallback: question and options are rendered in the image
if "question" in doc and "options" in doc: # original operation
question = construct_prompt(doc, post_prompt)
if config["metadata"]["interleaved_format"]:
Expand Down
Loading