Partial Combinatorial Generation #762
-
Let's say I have the following:
Using combinatorial generation and a fixed seed, I would get 4 results: But now let's say I also want to pick a random artist from a wildcard file containing hundreds of artists:
Using combinatorial generation here will result in hundreds and hundreds of generations. Is it possible to get this kind of behavour? Note that there are also two ways this could go - either a random artist is selected for each of the 4 generations, e.g. black cat, by ??artist?? Or the random artist is picked first and the same one used each time black cat, by Bob Ross Ideally I'd be able to choose either option as both have their uses. |
Beta Was this translation helpful? Give feedback.
Replies: 3 comments 2 replies
-
Based on the samplers syntax you should be able to do
(tilde within the wildcard wrap) to override the sampler for that command. |
Beta Was this translation helpful? Give feedback.
-
I'm thinking this is impossible, but any idea how I might use combinatorial wildcards together with Prompt S/R on XYZ plot? The idea is to pick from the list of artists from an artist.txt wildcard file, and actually show that artist's name on the XYZ plot. |
Beta Was this translation helpful? Give feedback.
-
I thought as much. I guess prompt S/R is not really the right tool for the job anyway, as it doesn't even make sense to search and replace I guess the best I can do is just convert the wildcard list of artists into a comma separated line of artists, and use that instead in the S/R plot. Just means I won't be able to have any randomness, but this would'nt be needed anyway if I just want to compare all artists. |
Beta Was this translation helpful? Give feedback.
Based on the samplers syntax you should be able to do
(tilde within the wildcard wrap) to override the sampler for that command.