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
Your elemental merger is fantastic, but it changes the paradigm completely, in a good way. Roll the dice!
Typically I find using a low range 0 to 0.1 and multiple iterations, saving out the checkpoint and reselecting the new one.
To make it easier I have some suggestions:
I would like the random values to be from a normal distribution, where both peak value and width are parameters. This would make more sparse changes.
It would be even better if you could apply a gradA, gradV etc multiplier, effectively filtering the areas the elemental merge will affect.
For the workflow it would be great if you could set it to do 5 variants with 5 prompts 5 seeds on an xy grid then you easily choose the variant to be the new base, and repeat.
The text was updated successfully, but these errors were encountered:
Repository owner
locked and limited conversation to collaborators
Jan 18, 2025
Your elemental merger is fantastic, but it changes the paradigm completely, in a good way. Roll the dice!
Typically I find using a low range 0 to 0.1 and multiple iterations, saving out the checkpoint and reselecting the new one.
To make it easier I have some suggestions:
I would like the random values to be from a normal distribution, where both peak value and width are parameters. This would make more sparse changes.
It would be even better if you could apply a gradA, gradV etc multiplier, effectively filtering the areas the elemental merge will affect.
For the workflow it would be great if you could set it to do 5 variants with 5 prompts 5 seeds on an xy grid then you easily choose the variant to be the new base, and repeat.
The text was updated successfully, but these errors were encountered: