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.

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

seed is not fixed #2141

Closed
Alireza2317 opened this issue Feb 2, 2024 · 8 comments
Closed

seed is not fixed #2141

Alireza2317 opened this issue Feb 2, 2024 · 8 comments
Labels
wontfix / cantfix This will not be worked on

Comments

@Alireza2317
Copy link

Alireza2317 commented Feb 2, 2024

Hello there. i have a problem with fixed seeds.
I have created multiple images on random seed, then i liked the output with a certain seed like 5207638156172183669.

then i unchecked the random seed and entered the seed in the box, and created more outputs. but when i looked at the log html, the seeds were nor fixed nor completely random, they were increments of my main seed. i mean i created 4 outputs, the first one's seed was in fact 5207638156172183669, but the next one's was 5207638156172183670 (one more than the last), and the next was 5207638156172183671 and so on.

Of course the outputs were completely different due to the different seeds.

i tried other base models ( i tested on juggernutV6 and juggernutV8) and i tried unchecking all the styles but it happened again.

I don't exactly know if this is by design or not, but i just wanted to be sure what is going on.

Im using fooocus version 2.1.862 on Linux.

@mashb1t
Copy link
Collaborator

mashb1t commented Feb 3, 2024

The first image per generation is generated with the given seed, all subsequent images are increments on this seed. Increasing image amount with always fixed seed would cause all images to be the same, which is not what you'd expect.
Please decrease the image amount to 1 to always get the image with the exact seed you've set.

@mashb1t mashb1t closed this as not planned Won't fix, can't repro, duplicate, stale Feb 3, 2024
@mashb1t mashb1t added the wontfix / cantfix This will not be worked on label Feb 3, 2024
@Alireza2317
Copy link
Author

The first image per generation is generated with the given seed, all subsequent images are increments on this seed. Increasing image amount with always fixed seed would cause all images to be the same, which is not what you'd expect. Please decrease the image amount to 1 to always get the image with the exact seed you've set.

ok it makes sense. but what if I wanted to use wildcards?

@mashb1t
Copy link
Collaborator

mashb1t commented Feb 3, 2024

I don't know if i got what you mean but wildcards are random in every case.
Please check if #1761 or #1503 suit your needs.

@poor7
Copy link

poor7 commented Feb 3, 2024

ok it makes sense. but what if I wanted to use wildcards?

I think it's possible to use the "infinite generation" mode. Enabled and disabled by right-clicking on the Generate button.

@Alireza2317
Copy link
Author

I don't know if i got what you mean but wildcards are random in every case. Please check if #1761 or #1503 suit your needs.

I meant i wanna generate for example 10 images, using a __color__ wildcard, for the same seed. so every little detail is the same, except the color of something in the picture. but if that's not possible then i guess i should do this manually.

@mashb1t
Copy link
Collaborator

mashb1t commented Feb 3, 2024

ok it makes sense. but what if I wanted to use wildcards?

I think it's possible to use the "infinite generation" mode. Enabled and disabled by right-clicking on the Generate button.

@Alireza2317 this is then also the solution i'd propose, just right click the generate button and use genrate forever

@flannerybh
Copy link
Contributor

I don't know if i got what you mean but wildcards are random in every case. Please check if #1761 or #1503 suit your needs.

I meant i wanna generate for example 10 images, using a __color__ wildcard, for the same seed. so every little detail is the same, except the color of something in the picture. but if that's not possible then i guess i should do this manually.

Sounds like #1503 is exactly what you want. The only thing is you'd just type out the different colors in the array.

@Alireza2317
Copy link
Author

I don't know if i got what you mean but wildcards are random in every case. Please check if #1761 or #1503 suit your needs.

I meant i wanna generate for example 10 images, using a __color__ wildcard, for the same seed. so every little detail is the same, except the color of something in the picture. but if that's not possible then i guess i should do this manually.

Sounds like #1503 is exactly what you want. The only thing is you'd just type out the different colors in the array.

oh thank you so much, i did not know about arrays in the prpmpt.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
wontfix / cantfix This will not be worked on
Projects
None yet
Development

No branches or pull requests

4 participants