fix: adjust timestep calculations for DDIM and TCD #665
Conversation
This file contains hidden or 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
On img2img, the number of steps correspond to the last precalculated sigma values, but the internal alphas_cumprod and compvis_sigmas were being computed over the entire step range. Also, tweaks the prev_timestep calculation on DDIM to better match the current timestamp (like on TCD), to avoid inconsistencies due to rounding.
wbruna
force-pushed
the
fix_ddim_tcd_timesteps
branch
from
7aa4d93 to
2e0c46a
Compare
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Outdated suggestions cannot be applied.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.
On img2img, the number of steps correspond to the last precalculated sigma values, but the internal alphas_cumprod and compvis_sigmas were being computed over the entire step range.
Also, tweaks the prev_timestep calculation on DDIM to better match the current timestamp (like on TCD), to avoid inconsistencies due to rounding.
This partially fixes #663 , and seems enough to at least generate coherent images. But low strength generations still produce noisy outputs.