BAD INFLUENCE: MIX, MATCH & MUTATE

I've made a lot of LoRAs.

I've tested them, abused them, mixed them with each other, thrown them at different subjects, and generally done all the things you're supposed to do before releasing one into the wild.

But there's one thing I can't realistically test:

What happens when other people start mixing them with models I would never think to use? So I'm outsourcing the problem (because I'm terribly lazy).

BAD INFLUENCE -> LOCATED HERE

I'm launching a two-week bounty called Bad Influence: Mix, Match & Mutate, with 6,500 Gold Buzz up for grabs.

The idea is ridiculously simple:

Take any LoRA I've created and combine it with somebody else's work (or your own!)

Another style LoRA. A character. A concept. A completely different checkpoint. Three LoRAs stacked together in an arrangement that would make sensible people uncomfortable.

Whatever your weird little heart desires.

I want to see what happens when the models start influencing each other.

And my LoRA absolutely does not have to dominate the result.

Use it at 1.0. Use it at 0.3. Let another style take over. Throw it onto a checkpoint it was never intended for. I'm much more interested in an unexpected interaction than seeing somebody perfectly reproduce one of my model galleries.

6,500 GOLD BUZZ POOL!

Get weird. (I DEMAND IT) WEIRD FOR THE WEIRD GOD.

I don't care whether you're using one of my photography LoRAs, anime styles, illustration models, experimental garbage from the Junk Drawer, or something else entirely.

Mix genres. Cross ecosystems. Change weights. Make terrible decisions and see whether they produce excellent images. Because after making all these things, I've developed a pretty good idea of what they do when used correctly.

Now I'm considerably more interested in what happens when they're given bad influences.

█▓▒▒░░░I love you now gimmie a kiss░░░▒▒▓█