Skip to content

[Feature] MTP/Speculative Decoding - Disable model vision option #5534

Description

@Carl-P4

Testing MTP/Speculative Decoding with latest LLAMA.cpp shows that with a vision model like Qwen3.6-27B-MTP There is no gain in token generation with the vision mmproj loaded. If you don't enable vision by not passing the mmproj then you benefit from the MTP/Speculative decoding boost.

Ideally a way create a profile before loading a model or a way via the side menu to disable vision when using --spec-type draft-mtp

Metadata

Metadata

Assignees

No one assigned

    Labels

    feature requestFeature request pending on roadmap

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions