-
-
Notifications
You must be signed in to change notification settings - Fork 204
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.
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
Context Seed? #470
Comments
I can suggest another technique that might work with some tweaks: Use las last image of the last batch and force it as the first image of the next batch. You can use controlnet to force it but there is actually different ways depending of your workflow. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Hi, i am trying to understand whether it is possible to determine something like the context seed or context position , in order to make sure that concecutive batches are using the same context in a vid2vid workflow, so that they can be blended seamlessly..
i am also willing to dig in the code, but thought i ask first whether someone knows if thats even possible.
Particularly I want to address the problem of limited batch length, because of memory issues. So in order to process a longer video seamlessley, we would need to blend multiple batches together. And for that I am hoping to find out a way of how the last context of a batch can be the first context of the next batch, when taking into account overlap frames.
If you could point me in the right direction, that would be amazing :)
The text was updated successfully, but these errors were encountered: