Skip to content

Question about randomly sampled x' #3

@Liyw979

Description

@Liyw979

image

Is this randomly sampled x' nessesary?

Have you try using half of the x as student and another half of x as teaher for cs-kd loss and using the whole x for cross-entropy loss without introducing x'?Will this do bad to accuracy?

Thanks!

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