Skip to content

FullConv2d should accept input tensor to infer output size #102

@fmassa

Description

@fmassa

Having to specify adjW and adjH is not super handy. If the input size changes, those values might not be adapted anymore. One solution is to pass the desired output shape as input, as in nn.

Also, the order in out_padh and out_padw is doubly inverted here and here, it should be instead out_padw,out_padh.

I can send a PR if you agree with the changes.

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