Add seed resizing to community pipelines #1011
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
|
The documentation is not available anymore as the PR was closed or merged. |
|
Very cool! |
PhaneeshB
pushed a commit
to nod-ai/diffusers
that referenced
this pull request
Mar 1, 2023
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Outdated suggestions cannot be applied.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.
Part of #875 and bigger story #841
The code is 1:1 with core stable_diffusion pipeline, the only difference starting around line 265 in the pipeline file.
In order to achieve seed stability, we simply generate the original 512 x 512 seed, generate a second seed of arbitrary size, and then paste the first seed into the second seed. This yields greater stability.
As an example, generating: "A painting of a futuristic cop", we get
Reference image, 512 x 512:
Stable seed image, 512 x 592:
Non stable seed image, 512 x 592:
So! Not exactly / 100% the same (though automatic111 did not claim it to be) but certainly seems much more stable with this approach.