`convole2NN`: Why can't 0 padding be used? · Issue #2892 · arrayfire/arrayfire · GitHub
Skip to content

convole2NN: Why can't 0 padding be used? #2892

@JonathanWoollett-Light

Description

@JonathanWoollett-Light

The function convole2NN does not allow for any of the of the values of the padding parameter to be 0.

It is relatively common practice to use 0 padding in convolutional layers.

Is this an oversight or a decision? If a decision, why?

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