Skip to content

Clarification on QueryType.question_image_and_text in generate_prompts.py #19

Description

@Loyal-Young

Hi~~,thank you for open-sourcing this helpful work.

I was looking through generate_prompts.py and noticed that gen_query() includes the following branch:
elif query_type == QueryType.question_image_and_text:
return "Understand the question in the image and answer it.", text_to_image(
wrap_text(question)
)
However, QueryType currently seems to define only:
class QueryType(IntEnum):
prompt_6 = 2
figstep = 6
baseline = 7
prompt_5 = 8

I may be missing some context, but it looks like QueryType.question_image_and_text is not defined in the enum. As a result, calling gen_query() with query types such as figstep, baseline, or prompt_5 may raise an AttributeError before reaching their intended branches.
Could you clarify whether question_image_and_text was intentionally omitted, or whether it should be added back to QueryType? If it was used in an earlier experiment setting, it would also be helpful to know what enum value it should have.
Thanks again for releasing the code and dataset.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions