Skip to content

ConvTranspose3d does not accept non-uniform kernelSize/stride values #1274

New issue
New issue
Closed
Closed
ConvTranspose3d does not accept non-uniform kernelSize/stride values#1274

Description

@rogerOxHB
rogerOxHB
opened on Mar 22, 2024

I am trying to replicate a pytorch UNet model where one of the layers is defined as

ConvTranspose3d(320, 320, kernel_size=(1, 2, 2), stride=(1, 2, 2))

TorchSharp does not expose an interface to specify kernel sizes and strides as a tuple, but only as single long values.

The same is true for ConvTranspose2d.

Activity

  1. NiklasGustafsson
    added a commit that references this issue on Mar 22, 2024

    Merge pull request #1275 from NiklasGustafsson/missing

    db565be

  2. NiklasGustafsson commented on Mar 22, 2024

    @NiklasGustafsson
    NiklasGustafsson
    on Mar 22, 2024

    PR merged. Release coming up.

  3. NiklasGustafsson commented on Mar 26, 2024

    @NiklasGustafsson
    NiklasGustafsson
    on Mar 26, 2024

    v0.102.3 was just released with the fix for this.

  4. NiklasGustafsson
    closed this as completedon Mar 26, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

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