Refactor Torch-space upscale fully out of ScuNET/SwinIR #14484
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
akx
force-pushed
the
swinir-resample-for-div8
branch
2 times, most recently
from
ca93008 to
30cac6a
Compare
akx
force-pushed
the
swinir-resample-for-div8
branch
from
30cac6a to
cf14a6a
Compare
AUTOMATIC1111
approved these changes
Jan 2, 2024
Closed
Closed
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Continues where #14476 left off. Please ignore the branch name, this got on a tangent.
This PR refactors the PIL-to-tensor-and-back code from the ScuNET/SwinIR into more reusable helpers.
Next we'll probably want to see whether to use the Torch tiling code or the PIL tiling code.
Screenshots/videos:
No user-facing changes. Again, this works on my Macbook, but I think there could be some device/dtype strangeness I'd appreciate other people to poke at too 😁
Checklist: