You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
recently I tried the mobile app "Stable Horde Flutter". I tried to replicate the results in A1111, but there is a strong deviation. So I thought it might has to do with the app itself and I installed the extension for "Stable Horde".
There is a strong difference between the images. Please check my results:
This is the prompt:
(car, technical, copper, metal, rust, tubes, screws, detailed), (outdoor), ((steam+++)), (sparks++), coal++, (cinematic lighting), banana
banana is an example, as it shows the problem much better.
V1 is A1111 on my local PC. V2 is A1111 on my local PC with the setting "Emphasis: use (text) to make model pay more attention to text and [text] to make it pay less attention" disabled.
"Stable Horde" is the result I get via the extension.
The image I get from the "Stable Horde" extension seems odd. The car is gone and it looks like the word banana has a big influence on the resulting image, which just looks jittered.
I assume it has to do with how the prompt weights are handled. Model is the same, Sampler, seed and cfg is the same as well.
Is there a way to approximate the results from the extension and A1111 from my local PC?
I would like to get a similar result from A1111, that I get via the "Stable Horde" extension.
And I also would like to get similar results from the "Stable Horde" extension that I get from A1111 locally.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
-
Hi,
recently I tried the mobile app "Stable Horde Flutter". I tried to replicate the results in A1111, but there is a strong deviation. So I thought it might has to do with the app itself and I installed the extension for "Stable Horde".
There is a strong difference between the images. Please check my results:
This is the prompt:
(car, technical, copper, metal, rust, tubes, screws, detailed), (outdoor), ((steam+++)), (sparks++), coal++, (cinematic lighting), banana
banana is an example, as it shows the problem much better.
V1 is A1111 on my local PC. V2 is A1111 on my local PC with the setting "Emphasis: use (text) to make model pay more attention to text and [text] to make it pay less attention" disabled.
"Stable Horde" is the result I get via the extension.
The image I get from the "Stable Horde" extension seems odd. The car is gone and it looks like the word banana has a big influence on the resulting image, which just looks jittered.
I assume it has to do with how the prompt weights are handled. Model is the same, Sampler, seed and cfg is the same as well.
Is there a way to approximate the results from the extension and A1111 from my local PC?
I would like to get a similar result from A1111, that I get via the "Stable Horde" extension.
And I also would like to get similar results from the "Stable Horde" extension that I get from A1111 locally.
Does anyone have an idea?
Beta Was this translation helpful? Give feedback.
All reactions