Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Discrepancies between Chesapeake datamodules/trainers #337

Closed
adamjstewart opened this issue Dec 30, 2021 · 0 comments · Fixed by #992
Closed

Discrepancies between Chesapeake datamodules/trainers #337

adamjstewart opened this issue Dec 30, 2021 · 0 comments · Fixed by #992
Labels
datamodules PyTorch Lightning datamodules trainers PyTorch Lightning trainers

Comments

@adamjstewart
Copy link
Collaborator

During #329, I noticed a lot of issues with our Chesapeake datamodules and how they integrate with BYOLTask and SemanticSegmentationTask. For example:

We should specify which dimension to squeeze over to fix the "batch size of 1 doesn't work" issue, and then fix the other issues that pop up when we do that. I believe there are some inconsistencies between:

  • the data returned by ChesapeakeCVPRDataModule when using prior/no priors
  • the data expected by SemanticSegmentationTask/BYOLTask
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
datamodules PyTorch Lightning datamodules trainers PyTorch Lightning trainers
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

DataModules: run all data augmentation on the GPU microsoft/torchgeo
1 participant
@adamjstewart