Skip to content

Video MME Reproduce (Question about think mode usage) #39

Description

@LimGeunTaekk

Hello, thank you for your great work.

I am currently working on reproducing the results for LongViLa-R1 7B on the VideoMME benchmark.

I noticed that the paper reports a score of 65.1 on VideoMME (without subtitles). Could you please clarify if this performance was achieved using "Think Mode"?

I would appreciate it if you could share the specific inference settings used to achieve this score.

Thank you in advance!

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions